{"as_of":"2026-08-17T01:05:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:50636911a5baae1845fdde9c6344c7b3ba8957c6e436f98b476446bde8c2d8f8","coverage":[{"denominator":73,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":73,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-16T12:19:04.253915Z","state":"measured"},{"denominator":74,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":74,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-16T06:30:59.297886+00:00","state":"measured"},{"denominator":1,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":1,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-02T21:28:29.665918Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.13558","snapshot_observed_at":"2026-08-02T21:28:29.665918Z","title":"Transformers can overcome the curse of dimensionality: A theoretical study from an approximation perspective","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2602.20555","last_updated":"2026-07-28T04:45:18Z","snapshot_observed_at":"2026-08-16T10:54:23.387763Z","submitted_at":"2026-02-24T05:14:01Z","title":"Standard Transformers Achieve the Minimax Rate in Nonparametric Regression with $C^{s,\\lambda}$ Targets","version":2},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-02T21:28:29.665918Z"},"links":{"cited_paper":"/paper/2504.13558","citing_paper":"/paper/2602.20555"},"observation_digest":"sha256:00525fb6fcdfb9849c83b9ac01ce69bfc10f0d13ccb261069376c25e0c9f05ef","observation_id":"53c04e93-8632-4176-9f9f-cd0403ff7dca","resolution":{"observed_at":"2026-08-02T21:28:29.665918Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/2504.13558/citation-record","integrity":"/paper/2504.13558/integrity","json":"/paper/2504.13558/citation-record.json","paper":"/paper/2504.13558"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.560673Z","title":"Vatt: Transformers for multimodal self-sup ervised learning from raw video, audio and text","venue":null,"work_id":"c9af4ce8-f740-43be-8572-8fc9decb7a59","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.795656Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:1d396594ebb1fb0ecbb507cbe1a4236d233d1e48cc53a2a33c38d0d37f7ccb51","observation_id":"42a8e14f-cf14-4459-8e02-51a8a3da8fe8","resolution":{"observed_at":"2026-08-16T12:19:05.565595Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.545822Z","title":"On the theory of dynamic programming","venue":null,"work_id":"00fd09a5-05e2-4551-b7c7-0f7e286bcf94","year":1952},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.801692Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:319aa1ee58727c755b3aa4a46313023542f7ff927b1c45e50b10f57d605b7dce","observation_id":"c7b1f536-ab6a-4de3-ac1f-9742322b8982","resolution":{"observed_at":"2026-08-16T12:19:05.550535Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.531153Z","title":"Is space -time attention all you need for video understanding? In International Conference on Machine Learning","venue":null,"work_id":"668932ef-8dca-4a53-a634-4b89292af5a2","year":null},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.806569Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:aa4c04599b15e7dffd41927893862af867ccecf6e36ab7f75f39b739c7e3b20b","observation_id":"0e2fc26b-7314-4cc9-943e-15888ee2deec","resolution":{"observed_at":"2026-08-16T12:19:05.536081Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.514884Z","title":"Simplicity bias in transformers and their ability to learn sparse boolean functio ns","venue":null,"work_id":"1a4520bd-9c25-496a-ad32-26274748a819","year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.811791Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:ffe479f3ff798886e5cc3c289d45213fdc0d4b52652fc2ecc1896e9b541bbded","observation_id":"c449ebf1-924c-4086-b342-222c50eb8761","resolution":{"observed_at":"2026-08-16T12:19:05.520139Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.499418Z","title":"Low-rank bottleneck in multi-head attention models","venue":null,"work_id":"34d70c81-9cf8-4ddc-82f0-1bce49bb11b9","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.816616Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:aea826040e1d50d2f96b62e5f1eaafe0cf192eb2a67668e1327960071de8c273","observation_id":"53127ccb-ee95-412b-84ad-8a4b2126e272","resolution":{"observed_at":"2026-08-16T12:19:05.504548Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.483766Z","title":"Language models are few-shot learners","venue":null,"work_id":"ad1a75a6-a3fe-4708-aa7c-11a8f68f8b98","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.821490Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:41e1c9c4e2cbe7f21199ef4b1b8c2a97051f5f2126bf80aa9c6022b8f49db144","observation_id":"9af09060-43f6-4d9a-ae00-c0db27aa3538","resolution":{"observed_at":"2026-08-16T12:19:05.488941Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.467766Z","title":"Decision trans former: Re- inforcement learning via sequence modeling","venue":null,"work_id":"410607bd-56f7-447c-977d-50ec57910939","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.826803Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:34078819c47285f1d7c3fabf4f97bf8f783d4c7c28f92dcc3d735cf2dec4beb8","observation_id":"03b89a9b-e749-487c-8af9-b5b69f4c8ec9","resolution":{"observed_at":"2026-08-16T12:19:05.472685Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.452762Z","title":"What can transformer learn with vary ing depth? case studies on sequence learning tasks","venue":null,"work_id":"8175f5b8-d61b-433d-83ee-4c55964005b3","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.831928Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:83c819dbfed1017699477ac84b4c3931887e852304236c8cb896c3ef5e6f9542","observation_id":"36a14d93-3ca0-4d6b-bdfe-9aaeab33d4e2","resolution":{"observed_at":"2026-08-16T12:19:05.457649Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.437548Z","title":"Approximation by superpositions of a sigmoidal function","venue":null,"work_id":"0ceffa70-29d4-48c3-bcd8-4fcfe2b65379","year":1989},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.836403Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:9ca3043e41fdc1f1b544480b035c97165b2e1e233ed4d19da20968316350e701","observation_id":"043d153b-bbf1-48da-8a15-f4819acb962b","resolution":{"observed_at":"2026-08-16T12:19:05.442478Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.422469Z","title":"BERT: pre- training of deep bidirectional transformers for language underst anding","venue":null,"work_id":"f11f2de7-2556-4e5d-a9af-f0f643b01d1d","year":2019},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.840916Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:8991d41f8b70dd060f9b24940695c6c437242a26b7eadfc809e7ab2ac01ded71","observation_id":"7a446817-6d96-428a-8bbc-2a0d0690cdbb","resolution":{"observed_at":"2026-08-16T12:19:05.427448Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.406258Z","title":"A ttention is not all you need: Pure attention loses rank doubly exponentially with depth","venue":null,"work_id":"33a26bc9-267a-4905-b8f0-6497a6141b1e","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.845484Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f751b437a2a40816b5e086bf0ceb4aa4d762ec482d6c6227e71da574dafec4dd","observation_id":"50bd2a6e-c50a-4501-bd35-e99c02fbfbbd","resolution":{"observed_at":"2026-08-16T12:19:05.411023Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.390835Z","title":"An image is worth 16x 16 words: Transformers for image recognition at scale","venue":null,"work_id":"c8483b7d-d31f-455c-8db0-f9bfdf59bdcd","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.850232Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:fe8ed221a345f95a9bd5fc35c0b90776fe7e013620bf88c3f4b1fa33094a9dd4","observation_id":"9d1ebcc3-84f2-442b-b5e7-78dad3beca95","resolution":{"observed_at":"2026-08-16T12:19:05.395944Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.375299Z","title":"Inductive biases and variable creation in self-attention mechanisms","venue":null,"work_id":"a3a8caa1-5943-4354-a08d-8d14408ab5ff","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.854731Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:3203b4f7f26c4ffa7dbaffe3f0275346c4f2ddc1db80997d9b682df35dc196ad","observation_id":"3a15c8cc-748d-4363-b785-2c9e50ba0369","resolution":{"observed_at":"2026-08-16T12:19:05.379905Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2202.12166","last_updated":"2022-02-24T16:06:01Z","snapshot_observed_at":"2026-08-16T17:18:54.055873Z","submitted_at":"2022-02-24T16:06:01Z","title":"Attention Enables Zero Approximation Error","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2202.12166","snapshot_observed_at":"2026-08-16T12:19:03.859260Z","title":"Attention enables zero approximation error","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.859260Z"},"links":{"cited_paper":"/paper/2202.12166","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f80d653337513dd073ac17e530b03e9c3a9f33ac1b8e82a07ea98ca53b5d4f3b","observation_id":"4f067f10-02b2-4fb2-a21c-12789e4c645a","resolution":{"observed_at":"2026-08-16T12:19:03.859260Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.360745Z","title":"Switch transfor mers: Scaling to trillion parameter models with simple and efficient sparsity","venue":null,"work_id":"e0190fdb-1fcd-46b3-8d8d-88ee24d2367b","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.864144Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:b638c9da12d6eacb5f663846a13cf41324fc12130f5ef44a05b19e0a8d5463f4","observation_id":"84621f43-dc7f-4811-a4b7-5d47ef01e51b","resolution":{"observed_at":"2026-08-16T12:19:05.365515Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.345535Z","title":"On the approximate realization of continu ous mappings by neural networks","venue":null,"work_id":"bb45f08f-cdab-471d-9b9c-cd9405105ada","year":1989},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.868559Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:b2077bc56db879ddee566afcf3581aed7d180acfa65bf97cb6922e3596fd5b58","observation_id":"5e8ad6a6-59c6-492e-beb3-92747c8b9f8a","resolution":{"observed_at":"2026-08-16T12:19:05.350418Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.328493Z","title":"Approximation rates for neur al networks with encodable weights in smoothness spaces","venue":null,"work_id":"c9a07f7e-c872-4123-8d0b-fe7fcaeff532","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.873065Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:6a47777f7a50f000cc51d58c5893c34c23d0ca3ea03bb984a50d6ad1fb280cad","observation_id":"dccedb40-47e8-4b97-9991-f46434a4d798","resolution":{"observed_at":"2026-08-16T12:19:05.334546Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.312351Z","title":"Error bou nds for approxima- tions with deep relu neural networks in w s, p norms","venue":null,"work_id":"fd74db89-fe3f-4383-ad39-b3df22f79aa9","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.877313Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:6a9a66e845317a676251e0a022203a6eae77a12252000711cfc3407ec33d023a","observation_id":"78906e9d-6911-4c45-ba9f-44e3bccc5b53","resolution":{"observed_at":"2026-08-16T12:19:05.317474Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.296173Z","title":"On the rate of convergence of a classifier based on a transformer encoder","venue":null,"work_id":"2fa0fe3b-102a-4f97-b070-de20d43a78c5","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.882016Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:fc4bd54446a53dd9f77411038bbee0add8754aff0b14c546a6ca24fcf31a68e1","observation_id":"2751d893-2ed2-403e-adca-07a8212bf298","resolution":{"observed_at":"2026-08-16T12:19:05.301367Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2411.06646","last_updated":"2024-11-11T01:05:28Z","snapshot_observed_at":"2026-08-16T13:01:20.253184Z","submitted_at":"2024-11-11T01:05:28Z","title":"Understanding Scaling Laws with Statistical and Approximation Theory for Transformer Neural Networks on Intrinsically Low-dimensional Data","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2411.06646","snapshot_observed_at":"2026-08-16T12:19:03.886484Z","title":"Understanding scaling laws with stat istical and ap- proximation theory for transformer neural networks on intrinsic ally low-dimensional data","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.886484Z"},"links":{"cited_paper":"/paper/2411.06646","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:c9d883776f23313c3bfc0d4c1c86987b36e249013a94567c8fc58f89e473f189","observation_id":"f8424206-ee0c-4ff8-a156-49fe7d49a1cd","resolution":{"observed_at":"2026-08-16T12:19:03.886484Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.279814Z","title":"Approximation capabilities of multilayer feedforwar d networks","venue":null,"work_id":"0529819e-52d0-42bd-8ead-ab5ab98f1259","year":1991},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.891343Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:e9e6a37cbe2842eac572faf2e47f206da29dd14612e53d7a87182f3025d5cd12","observation_id":"3fc298b6-8d70-48cd-bef0-bb77e4eca956","resolution":{"observed_at":"2026-08-16T12:19:05.285473Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2411.16525","last_updated":"2025-06-05T23:04:39Z","snapshot_observed_at":"2026-08-13T12:09:17.275284Z","submitted_at":"2024-11-25T16:12:17Z","title":"Fundamental Limits of Prompt Tuning Transformers: Universality, Capacity and Efficiency","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2411.16525","snapshot_observed_at":"2026-08-16T12:19:03.895916Z","title":"Fundamental limits of prompt tuning transformers: Univer sality, capacity and efficiency","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.895916Z"},"links":{"cited_paper":"/paper/2411.16525","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:dda0621b79ccabd12f9d6c114a1c3df3d64ff1029fac898c9b26feb71bcad4a7","observation_id":"5ea2b1b0-2a12-4da1-9a0c-fb9d2257abc2","resolution":{"observed_at":"2026-08-16T12:19:03.895916Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.262815Z","title":"Offline reinforcem ent learning as one big sequence modeling problem","venue":null,"work_id":"7850c32b-cbd5-4b29-95d6-3dd5050908a2","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.900721Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f7e35422f9f27d42879b1efa6af071a4fb711c09243a915563ed4353fd0f8732","observation_id":"c8e847a8-7654-46f5-8104-3466c466c251","resolution":{"observed_at":"2026-08-16T12:19:05.268484Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.18475","last_updated":"2025-01-02T05:02:48Z","snapshot_observed_at":"2026-08-16T15:28:43.707646Z","submitted_at":"2023-05-29T10:56:36Z","title":"Approximation Rate of the Transformer Architecture for Sequence Modeling","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.18475","snapshot_observed_at":"2026-08-16T12:19:03.905108Z","title":"Approximation rate of the transfo rmer architecture for sequence modeling","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.905108Z"},"links":{"cited_paper":"/paper/2305.18475","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:97ef32ac45fa31a07d352d6ab3cc698fc5af98c227f297c9001fb3ac1303e9c2","observation_id":"66605c13-bde2-47d6-ae01-1e5f915b8d79","resolution":{"observed_at":"2026-08-16T12:19:03.905108Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.246478Z","title":"Deep neural networks with relu-sine-exponentia l activations break curse of dimensionality in approximation on h¨ older class","venue":null,"work_id":"61610337-e226-4387-8920-11bfcb373d0d","year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.910195Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:7afbc8494603d29743b67f0a683bf614bf4cde0b880cb6260e967b35e77fa149","observation_id":"306498dc-2755-418b-b630-e971668a00c8","resolution":{"observed_at":"2026-08-16T12:19:05.251699Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2404.02538","last_updated":"2024-04-28T10:10:33Z","snapshot_observed_at":"2026-08-16T14:04:02.867871Z","submitted_at":"2024-04-03T07:50:53Z","title":"Convergence Analysis of Flow Matching in Latent Space with Transformers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.02538","snapshot_observed_at":"2026-08-16T12:19:03.914613Z","title":"Convergenc e analysis of flow matching in latent space with transformers","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.914613Z"},"links":{"cited_paper":"/paper/2404.02538","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:fbf3771748b7aee9c7f29d663e8d740fadd97512043bd51c3f10843ca2fea4f1","observation_id":"a4db15cc-a43d-499f-b9c2-4c9983642b5e","resolution":{"observed_at":"2026-08-16T12:19:03.914613Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.230332Z","title":"Approximation bounds for transformed with application to regression","venue":null,"work_id":"cbbfac56-13bd-4584-a169-a86c53f6c83b","year":2025},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.919632Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:e7bf2589c6718c2f2a76bbe0604daa0b1996b94436957bc3c883f78cb83f244a","observation_id":"17bceddc-08c6-462c-ba89-4ea54cf3441d","resolution":{"observed_at":"2026-08-16T12:19:05.235269Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2409.17677","last_updated":"2025-02-27T08:19:55Z","snapshot_observed_at":"2026-08-16T13:15:12.076331Z","submitted_at":"2024-09-26T09:36:47Z","title":"On the Optimal Memorization Capacity of Transformers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.17677","snapshot_observed_at":"2026-08-16T12:19:03.924459Z","title":"Optimal memorization capacity of transformers","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.924459Z"},"links":{"cited_paper":"/paper/2409.17677","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:a80df2092cf0be91429362f3b5dfdfe16304e0cdd8568a845decb818eebe5e9a","observation_id":"0f00dc10-d926-4b7b-96cf-88f2ff87a5bf","resolution":{"observed_at":"2026-08-16T12:19:03.924459Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.213081Z","title":"Are transformers with one lay er self-attention using low-rank weight matrices universal approximators? In International Conference on Learning Representations, 2024","venue":null,"work_id":"249a5f97-94bb-4467-80af-f5592a1a0a12","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.929504Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f5ab3bfcfcca30094065e89f00ce055850da714c00d453f20f67d8a78071516d","observation_id":"659e5ade-2944-44b7-82f1-012d481c8bc4","resolution":{"observed_at":"2026-08-16T12:19:05.218433Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.197604Z","title":"The lipschitz constant of self-attention","venue":null,"work_id":"cc82b084-4bdc-45cb-870b-fb36c3081bf4","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.933777Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:8ac9ea591e35fddbdb62ad2482f55212af3d2eca1508f1c6443cc353fa1868bc","observation_id":"7ebd27f9-a4cf-4fef-90f2-fa6422894d50","resolution":{"observed_at":"2026-08-16T12:19:05.202650Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.182515Z","title":"Provable memor ization ca- pacity of transformers","venue":null,"work_id":"ff5ad49d-fa1c-40e2-a1a2-c482af28156b","year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.938469Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:e334f0f37840eb79be3fed383d6678a7eca009f401408011a241a7d49c3cf43c","observation_id":"69ce6d86-8b27-44f8-b013-34846a5a9513","resolution":{"observed_at":"2026-08-16T12:19:05.187338Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.166765Z","title":"On the representation of cont inuous functions of many variables by superposition of continuous functions of one var iable and addi- tion","venue":null,"work_id":"600ec4c4-7764-4e52-9dcb-6a3b6f5e1d10","year":1957},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.942918Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:27ca0f2f098fec11f18f38a07b3d7471168667b759c435501f894e7bb13e103b","observation_id":"e36a25e0-0a7f-4abd-99e2-677b95ac4c5d","resolution":{"observed_at":"2026-08-16T12:19:05.171881Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.151239Z","title":"Univer- sal approximation under constraints is possible with transformers","venue":null,"work_id":"bf5b053b-e108-4b21-9035-6cd32cd3bf67","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.947407Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:443de26aac397ecbbb40be51cf4d7db98bce46cee48a3fad628693f599a10f71","observation_id":"4d593d45-ffca-47d2-b9de-dd82dffc9298","resolution":{"observed_at":"2026-08-16T12:19:05.156279Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.135604Z","title":"On the expressive flexibility of self-attention matrices","venue":null,"work_id":"939591e3-bbb5-4011-a1ce-180ab1bded25","year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.952146Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f80bfa380f7ea0f4cf30ad6131a3c78770b816065c8f84b913d102584bff346c","observation_id":"65e5ec0f-1216-40c2-a966-503795507abc","resolution":{"observed_at":"2026-08-16T12:19:05.141099Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1907.11692","last_updated":"2019-07-26T17:48:29Z","snapshot_observed_at":"2026-08-16T14:33:50.657682Z","submitted_at":"2019-07-26T17:48:29Z","title":"RoBERTa: A Robustly Optimized BERT Pretraining Approach","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1907.11692","snapshot_observed_at":"2026-08-16T12:19:03.956635Z","title":"RoBERT a: A robustly optimized BERT pretraining approach","venue":null,"work_id":null,"year":1907},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:03.956635Z"},"links":{"cited_paper":"/paper/1907.11692","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:0b4df45f82fa71e6678be4546a2a7ddeb5ec84a29c94c0fdf97ebaca484f586d","observation_id":"06a1dd5f-f456-4856-b455-2eb9a5357aae","resolution":{"observed_at":"2026-08-16T12:19:03.956635Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.120659Z","title":"De ep network ap- proximation for smooth functions","venue":null,"work_id":"a3aa6bf4-68a6-489f-9e0d-4f8107c24095","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.071550Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:206c0b00cf40cb38349e013af2cd6b0d4d9cb3fac6306a968f314623bcb1c83b","observation_id":"a43486ff-4ab4-421e-8406-5e565cffe9db","resolution":{"observed_at":"2026-08-16T12:19:05.125488Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.104907Z","title":"The ex- pressive power of neural networks: A view from the width","venue":null,"work_id":"9b5e0275-1b77-4750-9494-179f8e5ddedb","year":2017},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.076716Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:4f3359665c324ad9d9a17db655a8723a66a9e6e844855b05b13c9f2be8c789d4","observation_id":"1f1bb54a-83d3-4aaa-9d7b-4fbdd48a3a4b","resolution":{"observed_at":"2026-08-16T12:19:05.109663Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.089826Z","title":"Your transformer may not be as powerful as you expect","venue":null,"work_id":"8e8a570f-2170-4544-b9c8-194738bbc79e","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.081436Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:ae01fea8f192c23148f7880d70261a422c0c55fe76c711aad22dd28edead5afd","observation_id":"ad30d567-9a6d-4ead-99c0-b0818acb6605","resolution":{"observed_at":"2026-08-16T12:19:05.094696Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.086362Z","title":"Upper and lower mem- ory capacity bounds of transformers for next-token prediction","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.086362Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:7f9bb1351310707769d95f16603b2bef22a836453fd9ca46c110910875f7256e","observation_id":"d337b82c-0af8-4c98-a23f-111c31353c77","resolution":{"observed_at":"2026-08-16T12:19:04.086362Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.075075Z","title":"Memor ization capacity of multi-head attention in transformers","venue":null,"work_id":"2dc4ac2f-5568-4ef6-8d4a-ea7b17cbe437","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.091153Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:84de30586a8bb99adc5fdf2aa4408302328b44b069c6fd353bfd193507c8089c","observation_id":"5913a62c-90d8-40cc-ac94-24ce3dcbef59","resolution":{"observed_at":"2026-08-16T12:19:05.079960Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.059951Z","title":"Stabilizing transformers for reinforcement learning","venue":null,"work_id":"7f836e5c-66cf-40b4-887b-8a404b8c49a8","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.095882Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:b225792df4823258b34cdf651ca6b7d1557243c83723176115e540283a5e5f82","observation_id":"229435b5-cebe-45c2-a243-2aabbd9273d9","resolution":{"observed_at":"2026-08-16T12:19:05.064581Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.044641Z","title":"Scalable diffusion models with transfor mers","venue":null,"work_id":"229cc4c8-6835-4158-bb60-7b4ade6c4a6b","year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.100835Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:329e27e9c76d7bfa6bc9cbd221cf8be130c8a2173a0a4118b124fb8d16e3b36a","observation_id":"6e36f816-5683-4b81-a8a3-16973904a6e6","resolution":{"observed_at":"2026-08-16T12:19:05.049332Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.029607Z","title":"Prompting a pre trained trans- former can be a universal approximator","venue":null,"work_id":"46d1778e-d10b-4656-811f-8d6e44400d81","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.105631Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:6f7d5757b63f11f7225feeeb159b6f749d0de55ff403c187e11547a9f14cb6dd","observation_id":"801c7c72-afe2-42a6-bd7d-caa951ac53d7","resolution":{"observed_at":"2026-08-16T12:19:05.034434Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:05.014211Z","title":"Sutsk ever","venue":null,"work_id":"3ec01028-6e84-428f-abff-23ab5200008e","year":2018},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.110541Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f1f9c460bbf5fb5a76c124582cdd3fec98cd6057f69d97e22448100c82642775","observation_id":"3277a5f3-ca96-47b1-bcc8-2ad9afeeae74","resolution":{"observed_at":"2026-08-16T12:19:05.019193Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.998907Z","title":"Language models are unsupervised multitask learners","venue":null,"work_id":"5fe3c263-6f77-4532-92af-f992b4586553","year":2019},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.115764Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:783c38a0fb7af09190349a94bc5c52d3fc85cf7ccedd2153ef694607b1a6946a","observation_id":"c62a2d22-b60d-4f27-8014-119e8454275c","resolution":{"observed_at":"2026-08-16T12:19:05.003958Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.983423Z","title":"Represe ntational strengths and limitations of transformers","venue":null,"work_id":"6f0f9594-590f-42da-af31-a8c86c54c39e","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.120592Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:32d9b70dfbcb242c5875e2dad315bb10e21422a0ee0c20d02116af9275c7c407","observation_id":"de094312-5076-4e20-87c7-59c271f7d8cc","resolution":{"observed_at":"2026-08-16T12:19:04.988315Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.967032Z","title":"The kolmogorov–arnold represent ation theorem revisited","venue":null,"work_id":"edb14299-651d-4122-805c-f8f21b637642","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.125151Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:2141105e6ed6d01f5432ec6e90be4f9006d9a3f75115691c93a6f1973f92e467","observation_id":"cee3019a-9293-4197-ac22-09b6b9cc9b9e","resolution":{"observed_at":"2026-08-16T12:19:04.972345Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.951002Z","title":"Deep network a pproximation char- acterized by number of neurons","venue":null,"work_id":"9ea94cb5-b4ba-4bec-96c8-caa163bdb3c4","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.129813Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:8a0a15f9450d6419c3db7622af65329b0cf0a72faaac37f0c5f0b88f201c66d3","observation_id":"ec6d68b1-967e-49cd-89c2-4a4f93ca93f7","resolution":{"observed_at":"2026-08-16T12:19:04.956635Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.935091Z","title":"Deep network w ith approximation error being reciprocal of width to power of square root of depth","venue":null,"work_id":"01640fc1-4aa7-4b39-8aba-3bbb32ec285d","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.134443Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:9ab2b399368ae45084b23b99f419ec44eddd85498a0be08ddf4392ee5a3a64ea","observation_id":"20bb7d9f-f9a7-4078-9d93-55c27adb4cf3","resolution":{"observed_at":"2026-08-16T12:19:04.940149Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.917888Z","title":"Neural networ k approximation: Three hidden layers are enough","venue":null,"work_id":"cb084208-f53f-4acc-8fe0-2f896fa4449a","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.138973Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:18eda2d86acafaaa81b769e9fe2107005ba1b476e1d1ff7e2a58f6443e952876","observation_id":"eb945735-2a54-48b7-8abf-88d6e8529536","resolution":{"observed_at":"2026-08-16T12:19:04.923400Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.900993Z","title":"Optimal approx imation rate of relu networks in terms of width and depth","venue":null,"work_id":"1014cb93-3e67-47da-9481-96e04ca5ef9c","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.143954Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:9a658b918a60b34a7a0d829972dd94f16cdd911d809c9f0a4227925870f5a385","observation_id":"198c7600-5d2d-481d-b621-c6e90a362200","resolution":{"observed_at":"2026-08-16T12:19:04.906104Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.884544Z","title":"Sharp bounds on the approx imation rates, metric entropy, and n-widths of shallow neural networks","venue":null,"work_id":"3332e5c0-d848-4659-93c9-852fd73e9524","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.148778Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:0f6b458aee3f00254a610b3067404d96f9b3f842a54e514b3a89da8dfd8ef3a5","observation_id":"0ab4e756-b007-4950-8f38-d49f2a5b3733","resolution":{"observed_at":"2026-08-16T12:19:04.889911Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.868752Z","title":"Adaptivity of deep relu network for learning in beso v and mixed smooth besov spaces: optimal rate and curse of dimensionality","venue":null,"work_id":"e2a577a2-3b51-4850-8755-1f9b22e2b179","year":2019},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.154063Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:dc29717e9b03ab2e2aed89e2e1e69cf6ca1dddfa96ce702d59b3fce5885239f2","observation_id":"0a0d7df0-a005-49cd-862f-718ac7b32302","resolution":{"observed_at":"2026-08-16T12:19:04.874019Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.852635Z","title":"Approximation and estimat ion ability of trans- formers for sequence-to-sequence functions with infinite dimens ional input","venue":null,"work_id":"c1fd5e18-3bc5-4831-b607-3861d2c8dcf3","year":2023},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.158749Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:4b01aa1b8734f0ff24686fbf5c9bd8edc92c7ab54e108f3d1367c716a1191f3b","observation_id":"bf6de93d-7d02-4bd3-99e2-08fc02b34eff","resolution":{"observed_at":"2026-08-16T12:19:04.858018Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.836517Z","title":"Sequence length independen t norm-based gener- alization bounds for transformers","venue":null,"work_id":"433784e2-6804-4e2d-bb5d-9e5bf2867ddd","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.163893Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:ff4e96d5bb2073bdadf83ce209852673cbf6d04a9484b06f24ab4209f936e99f","observation_id":"615463b2-5879-4925-aaee-d985aac6d416","resolution":{"observed_at":"2026-08-16T12:19:04.841564Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.168933Z","title":"Gomez, Lukasz Kaiser, and Illia Polosukhin","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.168933Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:cb43e1fdf03cf46cd31a6c2ca1c267ee53afdb19d0ea2b87a4e790c7227ad71c","observation_id":"74052132-839c-41ea-a55e-bd85921f705d","resolution":{"observed_at":"2026-08-16T12:19:04.168933Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2007.02876","last_updated":"2020-07-20T13:57:49Z","snapshot_observed_at":"2026-08-14T04:58:30.142431Z","submitted_at":"2020-07-06T16:42:24Z","title":"A Mathematical Theory of Attention","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2007.02876","snapshot_observed_at":"2026-08-16T12:19:04.173444Z","title":"A mathematical theory of attention","venue":null,"work_id":null,"year":2007},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.173444Z"},"links":{"cited_paper":"/paper/2007.02876","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:2f46535354bd206571a2884545cbc5678d14f83818d1cf58ec29c84e228b9b2b","observation_id":"1dbad732-fc51-40dc-9652-6dacad3f1092","resolution":{"observed_at":"2026-08-16T12:19:04.173444Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.00522","last_updated":"2024-10-30T08:47:36Z","snapshot_observed_at":"2026-08-16T14:22:33.404838Z","submitted_at":"2024-02-01T11:43:13Z","title":"Understanding the Expressive Power and Mechanisms of Transformer for Sequence Modeling","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.00522","snapshot_observed_at":"2026-08-16T12:19:04.178534Z","title":"Understanding the expressive power and mechanisms of transformer for sequence modeling","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.178534Z"},"links":{"cited_paper":"/paper/2402.00522","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:83cf73d3df2cac664b526ba99fc16659d2064bc27ebb3ff5e225cd04bb1b2948","observation_id":"ba3ab9b3-6dcd-47d4-a030-7f953a6b827b","resolution":{"observed_at":"2026-08-16T12:19:04.178534Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":"2407.09580","doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.390490Z","title":"Don’t fear peculiar activation functions: Eu af and beyond","venue":null,"work_id":"88379a41-05f7-43fe-ab96-6816666f3911","year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.184559Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:30c2f3ac0711db6ab75782b44ee4384373e1d91a77724a332333e1fea89f4f3e","observation_id":"b189ae16-3a17-464e-a952-661551978161","resolution":{"observed_at":"2026-08-16T12:19:04.402054Z","resolver_source":"raw_fallback","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2406.06893","last_updated":"2024-06-11T02:15:53Z","snapshot_observed_at":"2026-08-16T13:44:20.874292Z","submitted_at":"2024-06-11T02:15:53Z","title":"Transformers Provably Learn Sparse Token Selection While Fully-Connected Nets Cannot","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.06893","snapshot_observed_at":"2026-08-16T12:19:04.190498Z","title":"Transfo rmers prov- ably learn sparse token selection while fully-connected nets cannot","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.190498Z"},"links":{"cited_paper":"/paper/2406.06893","citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:0ed5b727620e73f96c79aa9c49c04dd68ebc481b0e2e2ac03bdb5dca611a3f53","observation_id":"43a8d2bd-0325-4494-8186-da3b05dcbd25","resolution":{"observed_at":"2026-08-16T12:19:04.190498Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.811413Z","title":"Statistically meaningful app roximation: a case study on approximating turing machines with transformers","venue":null,"work_id":"9a47dc6c-644f-43cf-a3bb-cc18426d1ed2","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.195415Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:12203ca1810f667a23d99370eaffbe482b888f8d1828ae20041bf8ad38b6a690","observation_id":"a05e3f37-be23-4845-ad90-fce1f30d088d","resolution":{"observed_at":"2026-08-16T12:19:04.816457Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.796176Z","title":"Whic h transformer architecture fits my data? a vocabulary bottleneck in self-attent ion","venue":null,"work_id":"2051b4c3-69dc-40fa-a64c-ac364631155d","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.199930Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:c5146d4906a527ce007bead3e822a38eddbfe38e6a67846b2f658d4768f6d19f","observation_id":"57349ba1-68cf-4356-a546-7641e56f9b57","resolution":{"observed_at":"2026-08-16T12:19:04.801118Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.780631Z","title":"Carbonell, Ruslan Salak hutdinov, and Quoc V","venue":null,"work_id":"9a923bc7-fa1c-4fe3-951e-a4376ec06cc4","year":2019},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.204448Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:92755e8912b3bea857464a9173c3b2ba63335e559bc8dbe0810178ab7bfab433","observation_id":"9ee46c1b-00a4-4b87-9df9-48eb2c16f4da","resolution":{"observed_at":"2026-08-16T12:19:04.786289Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.764803Z","title":"Error bounds for approximations with deep r elu networks","venue":null,"work_id":"035c3037-e111-40be-8714-516c3cdd0f3d","year":2017},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.209548Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:85104938c1e61f0ce0c587f96afd3aca261e7d4b71134aae9e617a588029438c","observation_id":"caa4cfcf-42a4-47fd-b917-383990be018f","resolution":{"observed_at":"2026-08-16T12:19:04.770148Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.749200Z","title":"Optimal approximation of continuous functio ns by very deep relu networks","venue":null,"work_id":"d7b34c42-9f71-494d-a995-e17eb592d93e","year":2018},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":65,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.214138Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:76b60e0b38ead925c86220a4e00564f9348e766e1d6bef300f312c6f37ab6ec5","observation_id":"be7643b8-1f3f-4b7e-ae12-1240ea4cc897","resolution":{"observed_at":"2026-08-16T12:19:04.754503Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.733118Z","title":"Elementary superexpressive activations","venue":null,"work_id":"92915ef7-4da6-4b8b-9ed1-f922fb0e6906","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":66,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.219058Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:fa1dc18247b66f9ecd1786dd499323f08efed9d08b93c7e6cb673b303a4b417c","observation_id":"f3347c92-8afa-4a9f-a9b9-5b788eecbee1","resolution":{"observed_at":"2026-08-16T12:19:04.738981Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.716728Z","title":"The phase diagram o f approximation rates for deep neural networks","venue":null,"work_id":"420e2f0c-0dc2-4110-abca-c5bb4c7fba20","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":67,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.224164Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:3845fb70e175eecc2a7e26f7057fc8fc2ca12cf910637c3b9d02dd2aeffea5df","observation_id":"19007ddd-4229-48d2-95c5-45619b3630c5","resolution":{"observed_at":"2026-08-16T12:19:04.722334Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.699836Z","title":"Do transformers really perform badly fo r graph repre- sentation? In Advances in Neural Information Processing Systems , 2021","venue":null,"work_id":"9c282530-5a97-4aa1-b697-946987f28bb0","year":2021},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":68,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.229116Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:85b2f72745845847ef19b9b703c803241a054d38f9aff2c4f8dfacaae6b13ddc","observation_id":"9edcb588-19b4-4753-a27b-e3ae4b73e43f","resolution":{"observed_at":"2026-08-16T12:19:04.705752Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.682951Z","title":"Reddi, and Sanjiv Kumar","venue":null,"work_id":"5d6eec19-1f69-4096-85d3-d3b5de8f8264","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":69,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.233799Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:9491f96d14dbf046dec3412fa20692c0fda82b9dfa293c554d7ba7472f4bf63d","observation_id":"2d642a86-509e-4c3c-80d0-6108feb628e5","resolution":{"observed_at":"2026-08-16T12:19:04.688423Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.667718Z","title":"O (n) connections are expressive enough : Universal ap- proximability of sparse transformers","venue":null,"work_id":"db0d69fc-e547-4d40-81ee-956cc846e438","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":70,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.238563Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:6ac24ad3cad7910bd088f7d39da51136c6bfae505f98025d5826cd6e0abfba72","observation_id":"856a2c67-5713-4123-b9cd-a85339088395","resolution":{"observed_at":"2026-08-16T12:19:04.672541Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.650768Z","title":"Big bird: Transformers for longer sequences","venue":null,"work_id":"d622cd85-5ffb-4fa1-9d32-ec411bd220ec","year":2020},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":71,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.243357Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:cd9470598246e410b4d985827d19ec26567cdc85ba203dc6877b399f770f3289","observation_id":"bba17ba0-1e4d-48cf-a49d-a1394fede61a","resolution":{"observed_at":"2026-08-16T12:19:04.655986Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.633321Z","title":"Deep network a pproximation: Achieving arbitrary accuracy with fixed number of neurons","venue":null,"work_id":"da6c0bf6-da9b-44b9-9ab2-2033961f206b","year":2022},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":72,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.248010Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:f7b4da1ece767848fd2f2f7f4340d1b83223c726b4a1e6c0dfb53a5ac72341b6","observation_id":"45055737-d76b-4a81-92b7-2391c4e29e95","resolution":{"observed_at":"2026-08-16T12:19:04.640237Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T12:19:04.616733Z","title":"The space Cm consists of all functions whose first m derivatives exist and are continuous, and C0 denotes the space of continuous functions","venue":null,"work_id":"939efbf9-5b6f-4596-85a6-f5fc407f1fcf","year":null},"citing_paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective","version":1},"reference_index":73,"source":"pdf_text","source_observed_at":"2026-08-16T12:19:04.253915Z"},"links":{"citing_paper":"/paper/2504.13558"},"observation_digest":"sha256:9d0fc8e7d79d058fa3c8c6d1ade955ffc44fe5980bf8a4eb5a8344e9af7f743f","observation_id":"8bf1d0da-acde-4aab-b299-3d681bc1e659","resolution":{"observed_at":"2026-08-16T12:19:04.622766Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2504.13558","last_updated":"2025-04-18T08:56:53Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-16T12:03:40.481807Z","submitted_at":"2025-04-18T08:56:53Z","title":"Transformers Can Overcome the Curse of Dimensionality: A Theoretical Study from an Approximation Perspective"},"reference_resolution":{"displayed":73,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":12,"verified_exact":1,"verified_fuzzy":60},"total_outbound_references":73},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"thesis":"As of 17 August 2026, this Paper Citation Record lists 73 of 73 outbound references and 1 inbound Pith citation observation for arXiv:2504.13558."}