{"as_of":"2026-08-09T19:54:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:2c0a5659ac10c372afc95781f76c8b288ecf98d5e51af84aa3046b5fa7f4a837","coverage":[{"denominator":57,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":57,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-07T13:23:38.545260Z","state":"measured"},{"denominator":58,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":58,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-09T06:31:02.800959+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":"2026-05-18T13:36:55.938673Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"arxiv_reference","source_observed_at":"2026-05-18T13:41:25.898497Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"cited_work":{"arxiv_id":"2505.21987","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2505.21987","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Ace: Exploring activation co- sine similarity and variance for accurate and calibration-efficient llm pruning.arXiv preprint arXiv:2505.21987","venue":null,"work_id":"82088d0b-2aa6-4816-a2e9-7d12a6f050b9","year":null},"citing_paper":{"arxiv_id":"2509.22166","last_updated":"2026-04-24T11:44:43Z","snapshot_observed_at":"2026-07-06T22:30:50.642382Z","submitted_at":"2025-09-26T10:27:55Z","title":"Motivating Next-Gen Accelerators with Flexible (N:M) Activation Sparsity via Benchmarking Lightweight Post-Training Sparsification Approaches","version":4},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-05-18T13:36:55.938673Z"},"links":{"cited_paper":"/paper/2505.21987","citing_paper":"/paper/2509.22166"},"observation_digest":"sha256:979fb9484b0782535fe14c697e08f5a55830aea0031451955bd5bf87f49ae5f8","observation_id":"63ce8a56-3f83-4c7d-8636-48bae66434ea","resolution":{"observed_at":"2026-05-18T13:41:25.901495Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2505.21987/citation-record","integrity":"/paper/2505.21987/integrity","json":"/paper/2505.21987/citation-record.json","paper":"/paper/2505.21987"},"outbound":[{"citation":{"cited_paper":{"arxiv_id":"1810.04805","last_updated":"2019-05-24T20:37:26Z","snapshot_observed_at":"2026-07-30T09:12:38.100527Z","submitted_at":"2018-10-11T00:50:01Z","title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1810.04805","snapshot_observed_at":"2026-08-07T13:23:33.301394Z","title":"Bert: Pre-training of deep bidirectional transformers for language understanding","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.301394Z"},"links":{"cited_paper":"/paper/1810.04805","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:135e130b5de16299965accfaccc1c19e6e62ba5234b4ad52354204e45f5ec6b6","observation_id":"e6671fd3-03a3-4491-b3b2-3bb243256e7a","resolution":{"observed_at":"2026-08-07T13:23:33.301394Z","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":"2026-08-07T13:23:44.824326Z","title":"Language models are few-shot learners","venue":null,"work_id":"b577a3de-5a24-4e5c-87a9-d8476e1d13d9","year":2020},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.394522Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:5808d6d26e7f518a64c852005fa2a4f7e1e29d9d50b1a912ce93c6bdbc3bceb9","observation_id":"c77c457e-ad91-45d6-99f7-66970e79983b","resolution":{"observed_at":"2026-08-07T13:23:44.891786Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2302.13971","last_updated":"2023-02-27T17:11:15Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-02-27T17:11:15Z","title":"LLaMA: Open and Efficient Foundation Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2302.13971","snapshot_observed_at":"2026-08-07T13:23:33.537359Z","title":"Llama: Open and efficient foundation language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.537359Z"},"links":{"cited_paper":"/paper/2302.13971","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:00032a03281ed6eb35b20d80f9f4cbcf55252dc7bf2d0401732b04aaa1744a61","observation_id":"d1cd6627-8cd8-4cdb-9bdc-9903b9057777","resolution":{"observed_at":"2026-08-07T13:23:33.537359Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1606.05250","last_updated":"2016-10-11T02:42:36Z","snapshot_observed_at":"2026-08-08T11:05:45.903773Z","submitted_at":"2016-06-16T16:36:00Z","title":"SQuAD: 100,000+ Questions for Machine Comprehension of Text","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1606.05250","snapshot_observed_at":"2026-08-07T13:23:33.643584Z","title":"Squad: 100,000+ questions for machine comprehension of text","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.643584Z"},"links":{"cited_paper":"/paper/1606.05250","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:d6ede272b45a831ccdc504a1eb2403c9da6344ecf744c858c89b30f6fd1aed6a","observation_id":"1c57b68d-6963-4e2f-8321-c02b36de036b","resolution":{"observed_at":"2026-08-07T13:23:33.643584Z","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":"2026-08-07T13:23:44.540175Z","title":"Retrieval-augmented generation for knowledge-intensive nlp tasks","venue":null,"work_id":"b4ef2fbe-1623-43e8-b4a4-23a274cb6121","year":2020},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.703487Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:f247ce8a2d84d793ad0783cb41fbfddf3b9b911fa5c43ca17a1cd37bdd1e39da","observation_id":"f2c21b31-5573-4dbc-b8c4-949d27745193","resolution":{"observed_at":"2026-08-07T13:23:44.638976Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:44.403141Z","title":"Recipes for building an open-domain chatbot","venue":null,"work_id":"62a704d9-cdc5-4bd0-83d9-265893bd5402","year":2021},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.790950Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:5f77e400bf7714799ad0e43981e177a3bc0a73c8667adeded52e83e823680b74","observation_id":"b5ff4d06-a9c1-4231-a457-bf5ab77f8299","resolution":{"observed_at":"2026-08-07T13:23:44.465810Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2107.03374","last_updated":"2021-07-14T17:16:02Z","snapshot_observed_at":"2026-08-08T11:58:24.516369Z","submitted_at":"2021-07-07T17:41:24Z","title":"Evaluating Large Language Models Trained on Code","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2107.03374","snapshot_observed_at":"2026-08-07T13:23:33.932096Z","title":"Evaluating large language models trained on code","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:33.932096Z"},"links":{"cited_paper":"/paper/2107.03374","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:fd8f997d6fe8e9e52195f01a1c08bdbc2edc31c258621a94316379571ad0e441","observation_id":"a4867b9e-a33b-40a3-bab7-1e26b431f615","resolution":{"observed_at":"2026-08-07T13:23:33.932096Z","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":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T13:23:34.048484Z","title":"Bert: Pre-training of deep bidirectional transformers for language understanding","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.048484Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:84c2fa059402a0962f3e72b26ca1bc94de5e97a37ca796ac9eecb75592f49538","observation_id":"a5731580-0739-4654-9adc-62917981e5ad","resolution":{"observed_at":"2026-08-07T13:23:34.048484Z","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":"2026-08-07T13:23:44.230585Z","title":"Llm inference performance engineering: Best practices","venue":null,"work_id":"516e870a-9750-4521-aee1-80915c1ad686","year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.158980Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:d5c6d63593c99b409321306d64d193ad02f8e75dc29af3ad24a16ffd6c59d071","observation_id":"b9c4333b-876d-4e4f-9b2e-aae690ea6109","resolution":{"observed_at":"2026-08-07T13:23:44.299840Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2503.04992","last_updated":"2025-06-01T20:48:29Z","snapshot_observed_at":"2026-08-07T17:23:27.860798Z","submitted_at":"2025-03-06T21:42:35Z","title":"Wanda++: Pruning Large Language Models via Regional Gradients","version":4},"cited_work":{"arxiv_id":"2503.04992","doi":null,"metadata_source":"pith","pith_arxiv_id":"2503.04992","snapshot_observed_at":"2026-08-07T13:23:39.057729Z","title":"Wanda++: Pruning Large Language Models via Regional Gradients","venue":"cs.LG","work_id":"1285825d-f72b-4e0d-9be5-074f9ea8cd7a","year":2025},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.229021Z"},"links":{"cited_paper":"/paper/2503.04992","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:3a0203f4d77c163c875800837b1f4873a2a22b79db1b244876135fd2d1bfe9eb","observation_id":"db0d1abe-a396-4129-9a3b-84e7c9344286","resolution":{"observed_at":"2026-08-07T13:23:39.155438Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2012.15701","last_updated":"2021-07-22T13:13:45Z","snapshot_observed_at":"2026-08-09T15:34:02.499941Z","submitted_at":"2020-12-31T16:34:54Z","title":"BinaryBERT: Pushing the Limit of BERT Quantization","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2012.15701","snapshot_observed_at":"2026-08-07T13:23:34.333989Z","title":"Binarybert: Pushing the limit of bert quantization","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.333989Z"},"links":{"cited_paper":"/paper/2012.15701","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:73de46f1b31b762e777bb8b1c22e2f560b4facceeff8f68cd0599bb29fc8992e","observation_id":"ab7e5339-1b94-4628-b4ee-99949b2a5618","resolution":{"observed_at":"2026-08-07T13:23:34.333989Z","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":"2026-08-07T13:23:44.048498Z","title":"Spdy: Accurate pruning with speedup guarantees","venue":null,"work_id":"ac69f50e-1476-42d3-89ee-43c9547db436","year":2022},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.443115Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:cb9201c5b77d6287088f97854f3811128227887a5092c52b09108efc5d5c9c43","observation_id":"b6a945b9-402b-43ef-b4dc-0942f4da75c7","resolution":{"observed_at":"2026-08-07T13:23:44.130225Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:43.928517Z","title":"Smoothquant: Accurate and efficient post-training quantization for large language models","venue":null,"work_id":"3e1fd680-f2a2-4f3f-9c63-9d02a3421a59","year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.499768Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:ebab388e6c0e5bc27ba24054ba2f0ac94e716be70e88a4b9c7b4a2698aa47f67","observation_id":"d0f34d07-9173-4d9c-b003-4fd464212ee5","resolution":{"observed_at":"2026-08-07T13:23:43.967527Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:34.556385Z","title":"Awq: Activation-aware weight quantization for on-device llm compression and acceleration","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.556385Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:4e2c2c6b36b7581095c7b2815fcb06b0d65eaeb1287e1afd2cb81e9505a0c589","observation_id":"a458bac1-cf27-4dcf-ba18-49d8ba5a64cc","resolution":{"observed_at":"2026-08-07T13:23:34.556385Z","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":"2026-08-07T13:23:43.817209Z","title":"Optimal brain surgeon and general network pruning","venue":null,"work_id":"293d4c36-c8cc-4a93-95d3-a8806aafb80e","year":1992},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.657145Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:2578ed3cad3710c802a7066e7d29a7be5fdd9619c1630c0d8991a3ff2c05aa5a","observation_id":"d0cde6ae-f46f-44db-94fb-25550c62a5ec","resolution":{"observed_at":"2026-08-07T13:23:43.863098Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:34.734835Z","title":"Optimal brain damage","venue":null,"work_id":null,"year":1989},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.734835Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:9c48351f03a15009ba9c3c20734083f16a78b5f1aa87239b257ea2b724ff81ed","observation_id":"abb6c715-dc40-4c35-a510-ec2f29cdd33a","resolution":{"observed_at":"2026-08-07T13:23:34.734835Z","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":"2026-08-07T13:23:43.670491Z","title":"Scalable training of artificial neural networks with adaptive sparse connectivity inspired by network science","venue":null,"work_id":"940923d6-c1a5-4456-beb0-523bf1a8d011","year":2018},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.855968Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:6f0bd0f813dd97c07b2b56ea19d7e863cdb17e2866a58465b65cefc6ce4ae277","observation_id":"b492bedb-b1a2-4185-bd9e-303f858336b2","resolution":{"observed_at":"2026-08-07T13:23:43.733489Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2306.11695","last_updated":"2024-05-06T17:47:01Z","snapshot_observed_at":"2026-07-06T15:44:42.776459Z","submitted_at":"2023-06-20T17:18:20Z","title":"A Simple and Effective Pruning Approach for Large Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2306.11695","snapshot_observed_at":"2026-08-07T13:23:34.927550Z","title":"A simple and effective pruning approach for large language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.927550Z"},"links":{"cited_paper":"/paper/2306.11695","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:e43b4079426c7128985aa71e37f90800b22924e51f73e00d7aa3c4f8a1d7aca1","observation_id":"2481574c-b90a-49ef-b89b-f20d20db50b1","resolution":{"observed_at":"2026-08-07T13:23:34.927550Z","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":"2026-08-07T13:23:43.544983Z","title":"Sparsegpt: Massive language models can be accurately pruned in one-shot","venue":null,"work_id":"7da02fb3-6873-4032-8536-e998f3360e3f","year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:34.996367Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:702fc2bd9b379cea9a5aef05904eb36c4ce3e277c8cca06eb540332730b51f97","observation_id":"95da3411-5cb3-4c2f-ba3e-d55e9a0c394d","resolution":{"observed_at":"2026-08-07T13:23:43.615542Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2207.00112","last_updated":"2022-06-30T21:57:07Z","snapshot_observed_at":"2026-07-06T13:26:38.135401Z","submitted_at":"2022-06-30T21:57:07Z","title":"Language model compression with weighted low-rank factorization","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2207.00112","snapshot_observed_at":"2026-08-07T13:23:35.080757Z","title":"Language model compression with weighted low-rank factorization","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.080757Z"},"links":{"cited_paper":"/paper/2207.00112","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:f9f6ece9175bbf8cbc0b1151032afeec30251767d7109b23128fd086e6664c47","observation_id":"520d8110-9129-4f9b-affb-c7bd169a1da9","resolution":{"observed_at":"2026-08-07T13:23:35.080757Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.11417","last_updated":"2024-02-18T01:20:00Z","snapshot_observed_at":"2026-08-09T15:30:20.791421Z","submitted_at":"2024-02-18T01:20:00Z","title":"LoRETTA: Low-Rank Economic Tensor-Train Adaptation for Ultra-Low-Parameter Fine-Tuning of Large Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.11417","snapshot_observed_at":"2026-08-07T13:23:35.169522Z","title":"Loretta: Low-rank economic tensor- train adaptation for ultra-low-parameter fine-tuning of large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.169522Z"},"links":{"cited_paper":"/paper/2402.11417","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:c82e7c9a0b9c0a2c67af449ecdfd2a9e25cf25b9f1aa3e8ff52c6b6d0995fb0e","observation_id":"9a932015-6f65-4061-84e5-e35e7bc2591b","resolution":{"observed_at":"2026-08-07T13:23:35.169522Z","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":"2026-08-07T13:23:43.403864Z","title":"How contextual are contextualized word representations? comparing the geometry of bert, elmo, and gpt-2 embeddings","venue":null,"work_id":"ee402da6-cb78-4b10-bad3-1ff951d720b1","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.222818Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:9e802632056fdba58430334542d75e211c8dcf3a25d38f7b57a46560ac047d09","observation_id":"c640ff9e-5aa4-4ac9-a493-4411f6de0616","resolution":{"observed_at":"2026-08-07T13:23:43.469259Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:43.251900Z","title":"On the degeneration of neural text generation","venue":null,"work_id":"0bf651e3-70b4-4ef8-b161-794cdee55471","year":2020},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.286552Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:a9c74213aa93a20eabcc394df2c004b667b4f9422e43622a186cd720045b924c","observation_id":"30934107-35f6-4307-bd9a-0c2ef0cb5385","resolution":{"observed_at":"2026-08-07T13:23:43.342956Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1301.3781","last_updated":"2013-09-07T00:30:40Z","snapshot_observed_at":"2026-07-06T03:04:11.148340Z","submitted_at":"2013-01-16T18:24:43Z","title":"Efficient Estimation of Word Representations in Vector Space","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1301.3781","snapshot_observed_at":"2026-08-07T13:23:35.361824Z","title":"Efficient estimation of word representations in vector space","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.361824Z"},"links":{"cited_paper":"/paper/1301.3781","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:b96095d3caacdfcb1e71695a332edd789e466bdf3d180087b3fa09fe5b057fb5","observation_id":"be9d1bf5-1503-4aa5-8f2a-376960621877","resolution":{"observed_at":"2026-08-07T13:23:35.361824Z","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":"2026-08-07T13:23:43.132652Z","title":"Simcse: Simple contrastive learning of sentence embeddings","venue":null,"work_id":"dcbd5d00-808d-4efe-b2df-7d1238b72834","year":2021},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.428014Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:bcbec8c731637e304b5790b98301ba5906728d2be2ce85825a9e4ed145f1b54c","observation_id":"6a36b714-8907-47ad-9af3-4565e249d45b","resolution":{"observed_at":"2026-08-07T13:23:43.176880Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1510.00149","last_updated":"2016-02-15T06:25:40Z","snapshot_observed_at":"2026-08-04T16:59:47.843960Z","submitted_at":"2015-10-01T09:03:44Z","title":"Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1510.00149","snapshot_observed_at":"2026-08-07T13:23:35.482646Z","title":"Deep compression: Compressing deep neural net- works with pruning, trained quantization and huffman coding","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.482646Z"},"links":{"cited_paper":"/paper/1510.00149","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:87f14564125cbfdf9ffaf9745f2223dba5d0f43d6eaae537b890dd9cdc5071b9","observation_id":"750b1373-5438-4920-bc0b-dd680296efd0","resolution":{"observed_at":"2026-08-07T13:23:35.482646Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1810.05270","last_updated":"2019-03-05T05:58:11Z","snapshot_observed_at":"2026-07-06T07:07:34.689643Z","submitted_at":"2018-10-11T22:15:28Z","title":"Rethinking the Value of Network Pruning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1810.05270","snapshot_observed_at":"2026-08-07T13:23:35.550328Z","title":"Rethinking the value of network pruning","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.550328Z"},"links":{"cited_paper":"/paper/1810.05270","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:b933ec760eb26f0840ff24a13950d0c31f784b9c33c21a05eab76377a4a3f599","observation_id":"9883219d-015a-4d0c-b846-b06cc954d000","resolution":{"observed_at":"2026-08-07T13:23:35.550328Z","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":"2026-08-07T13:23:42.987616Z","title":"Learning both weights and connections for efficient neural networks","venue":null,"work_id":"c2d9cc43-e88e-4db4-b313-a5da95ffc04d","year":2015},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.606635Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:6e7c5cc4d59afe3bf3a862b8336541ddf74d68ea8e2af86ee4b182f4ce6c7e37","observation_id":"651ccc87-5d3d-41e1-ae00-5edd27cdafdf","resolution":{"observed_at":"2026-08-07T13:23:43.055006Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:42.831034Z","title":"The lottery ticket hypothesis: Finding sparse, trainable neural networks","venue":null,"work_id":"ec2da195-be6d-4bf8-9bb5-67928c7e4fa5","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.702063Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:31041cbf5e53ca1ac9ed09ac17e9d2e49c46e85966dd284690d68d88f51c9c63","observation_id":"8326899e-6ec7-410b-a478-819bd31219ae","resolution":{"observed_at":"2026-08-07T13:23:42.903988Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:42.616555Z","title":"Learning efficient convolutional networks through network slimming","venue":null,"work_id":"76a906bc-9f1e-4381-95af-c9e8f638d70e","year":2017},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.768014Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:f335433912a7c8d75e16cbdbab431517f25c58d79b1a2a0bfad8ca8226268d4e","observation_id":"37ce33fb-0290-4efa-82a0-cfb3c6bcf029","resolution":{"observed_at":"2026-08-07T13:23:42.732619Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:42.474288Z","title":"Importance estimation for neural network pruning","venue":null,"work_id":"db15173a-863c-4fac-bada-5361a202839a","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.826843Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:37aa7b073c43cc6596814bd3b528c186debc5e07ddc89217718a9277484172b6","observation_id":"0cc6b097-8fbf-413a-b38d-1ad47b46450a","resolution":{"observed_at":"2026-08-07T13:23:42.530619Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:42.278321Z","title":"Snip: Single-shot network pruning based on connection sensitivity","venue":null,"work_id":"17c93c39-e022-4b82-b53e-dabde5dff849","year":2018},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.894798Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:5f5c2a80a4371bd22b59dc0fc2292c387d574ab637fc8e2b6da8e275e25d32e8","observation_id":"870fa892-fb91-4fb4-9a28-115e7c2a2dde","resolution":{"observed_at":"2026-08-07T13:23:42.365037Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:42.142471Z","title":"Pruning filters for efficient convnets","venue":null,"work_id":"f691ae17-a67d-4ebc-bcde-2b3e4da534f2","year":2017},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:35.957727Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:098d5f24c8037b1f3ca5276bf0752bc2a0cbb952779ae4744d1680e835c0e6c3","observation_id":"1188103e-fcc4-4095-8d21-05ef2cf2b8c4","resolution":{"observed_at":"2026-08-07T13:23:42.210155Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:41.902939Z","title":"Accelerating sparse deep neural networks","venue":null,"work_id":"8647e5e1-2906-4ce0-b67c-ee2787716d2d","year":2021},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.024094Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:8e6f630122a1fdd050e7bcb865f8776e55838ba304ca99e2382d85f7d4a8a5d3","observation_id":"93fd6604-e317-43a3-a0c9-5c9f6e7de7aa","resolution":{"observed_at":"2026-08-07T13:23:42.073777Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:41.795574Z","title":"Optimal brain compression: A framework for accurate post-training quantization and pruning","venue":null,"work_id":"1fcb18b7-f89f-43a9-ada0-2ed79ce3e3cc","year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.080144Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:e821fd5d8088820d0db0b51b862c58e7243f6d5e6e0b247b24118ef2c241f399","observation_id":"b912dab5-1633-467d-b605-ceddcd1128eb","resolution":{"observed_at":"2026-08-07T13:23:41.841071Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.07883","last_updated":"2024-09-11T13:56:14Z","snapshot_observed_at":"2026-08-09T16:43:36.328935Z","submitted_at":"2024-02-12T18:48:50Z","title":"Equivalence of cost concentration and gradient vanishing for quantum circuits: An elementary proof in the Riemannian formulation","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.07883","snapshot_observed_at":"2026-08-07T13:23:36.134983Z","title":"Plug- and-play: Hardware-aware semi-structured pruning for large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.134983Z"},"links":{"cited_paper":"/paper/2402.07883","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:12c9f9b5e8307907a863d6d7315d04d202df3d236f050efac726021a0c0daae5","observation_id":"efe63dbd-4e8d-4741-b4b4-c1e46235e1ee","resolution":{"observed_at":"2026-08-07T13:23:36.134983Z","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":"2026-08-07T13:23:41.643678Z","title":"The state of sparsity in deep neural networks","venue":null,"work_id":"0ce14786-29fb-453b-a2c0-807c3c2cc220","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.214197Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:b10295ae1d6a81dc77179ebe488257abe2b1134ea05f3549f2d34e2313033027","observation_id":"e8cdef6a-fc9e-4534-91b3-2767f02ff433","resolution":{"observed_at":"2026-08-07T13:23:41.717311Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2406.02924","last_updated":"2024-06-05T04:25:23Z","snapshot_observed_at":"2026-08-09T09:34:57.546210Z","submitted_at":"2024-06-05T04:25:23Z","title":"Pruner-Zero: Evolving Symbolic Pruning Metric from scratch for Large Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.02924","snapshot_observed_at":"2026-08-07T13:23:36.317334Z","title":"Pruner-zero: Evolving symbolic pruning metric from scratch for large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.317334Z"},"links":{"cited_paper":"/paper/2406.02924","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:99d0d7d23afaa6f4bb2d441d3a1f8b3b27bd2a09c3050502ab09bc24b2283ad0","observation_id":"c3debba7-4a69-4d9f-840e-ed4314728916","resolution":{"observed_at":"2026-08-07T13:23:36.317334Z","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":"2026-08-07T13:23:41.489547Z","title":"Wanda: Weight-norm based pruning for efficient large language models","venue":null,"work_id":"796a987d-8af3-409f-bc50-064cf2bd2212","year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.374748Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:ad47ca33e8d5bfa9f636b67bd6564ee905f58eda9c8309b8daa8848633f12bf6","observation_id":"ee00874f-32d8-4507-81eb-0344f0dcf71d","resolution":{"observed_at":"2026-08-07T13:23:41.566595Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2502.10735","last_updated":"2025-05-23T03:17:51Z","snapshot_observed_at":"2026-08-07T18:16:05.489705Z","submitted_at":"2025-02-15T09:17:38Z","title":"Beyond One-Size-Fits-All Pruning via Evolutionary Metric Search for Large Language Models","version":2},"cited_work":{"arxiv_id":"2502.10735","doi":null,"metadata_source":"pith","pith_arxiv_id":"2502.10735","snapshot_observed_at":"2026-08-07T13:23:38.718375Z","title":"Beyond One-Size-Fits-All Pruning via Evolutionary Metric Search for Large Language Models","venue":"cs.CL","work_id":"13db5876-74bc-46c3-a7d3-a234f63d6876","year":2025},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.464599Z"},"links":{"cited_paper":"/paper/2502.10735","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:e3ee26847153f228bc7461f0b119178f6b24144408f666d6bc20b6c1404f0858","observation_id":"9dee79b5-d804-4f85-8f70-2e42772f41c4","resolution":{"observed_at":"2026-08-07T13:23:38.800446Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:41.319012Z","title":"Compression of deep neural networks","venue":null,"work_id":"7b8577c8-1aec-471a-a7dd-7816b77bceea","year":2021},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.523805Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:e47ad698994b77495667dc59f3e796a1e2fc4f20e537e4b78480a2fa0790282b","observation_id":"c1422a1d-016c-48d4-afad-c8cab9f74450","resolution":{"observed_at":"2026-08-07T13:23:41.385193Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:41.159642Z","title":"Hawq: Hessian aware quantization of neural networks with mixed-precision","venue":null,"work_id":"1b952d5b-09d8-4e6b-b6ca-55b6c5aa27dd","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.635522Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:90fa7b2c3aba4d7ee1c1495e02467dff8683b0b893bdc5cdecc99addab75a365","observation_id":"7cfe0b06-5b6f-43e4-a7ee-6ba496314b4f","resolution":{"observed_at":"2026-08-07T13:23:41.232664Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:40.906840Z","title":"Distributed representations of words and phrases and their compositionality","venue":null,"work_id":"859c73fd-afb9-420c-9cdc-46cdc8d8c978","year":2013},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.784754Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:8a51702080d93c0c0920f8c0567fee481b0db7c23a1b3df8cb939836880f3c39","observation_id":"f5db0301-2b75-47d8-88bd-eb3ee2ca8deb","resolution":{"observed_at":"2026-08-07T13:23:41.020307Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:36.954305Z","title":"Glove: Global vectors for word representation","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:36.954305Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:f4996f5b5d14937f3b3756abed2d9299f62c30adf45975414d58be49ca245c9f","observation_id":"2fe3dae5-f691-49dc-9336-0201514f9af7","resolution":{"observed_at":"2026-08-07T13:23:36.954305Z","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":"2026-08-07T13:23:40.712298Z","title":"Analyzing and measuring bert’s under- standing of syntax","venue":null,"work_id":"c45e581d-832e-4739-b91f-0c3f2c479563","year":2020},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.058481Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:3923a6d8eb6d155f7affa7c85c3c8b9b4eb902fa25bf61c2ee258508aa0dc802","observation_id":"5ec92824-77b8-4049-850c-0995a66e6f6d","resolution":{"observed_at":"2026-08-07T13:23:40.802174Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:40.543986Z","title":"Optimal brain surgeon and general network pruning","venue":null,"work_id":"5da3c941-a5e0-4414-81c9-1e0e76e0eeed","year":1993},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.182053Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:4bb4175b26d1081d91e21570bb0bd961e753432386324be1b27abaeb786c5a5e","observation_id":"efbfc5a8-c937-43a9-8594-14e451326251","resolution":{"observed_at":"2026-08-07T13:23:40.603082Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2307.09288","last_updated":"2023-07-19T17:08:59Z","snapshot_observed_at":"2026-08-07T12:56:43.323460Z","submitted_at":"2023-07-18T14:31:57Z","title":"Llama 2: Open Foundation and Fine-Tuned Chat Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2307.09288","snapshot_observed_at":"2026-08-07T13:23:37.311829Z","title":"Llama 2: Open foundation and fine-tuned chat models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.311829Z"},"links":{"cited_paper":"/paper/2307.09288","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:9c2878f8356f61af467eae4c0a74ff9c12dc412c1abffbd8b453a14d63abe912","observation_id":"5a58d127-136e-4acc-a9d3-5e78d0960274","resolution":{"observed_at":"2026-08-07T13:23:37.311829Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2205.01068","last_updated":"2022-06-21T17:04:40Z","snapshot_observed_at":"2026-08-06T03:13:37.403059Z","submitted_at":"2022-05-02T17:49:50Z","title":"OPT: Open Pre-trained Transformer Language Models","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2205.01068","snapshot_observed_at":"2026-08-07T13:23:37.444536Z","title":"Opt: Open pre-trained transformer language models","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.444536Z"},"links":{"cited_paper":"/paper/2205.01068","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:dce1c1483a442d0166c4cc77e774fab160472215ec54b21082724454e10cf031","observation_id":"6f6f00ad-0c1e-4541-8210-71eb3dfd237c","resolution":{"observed_at":"2026-08-07T13:23:37.444536Z","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":"2026-08-07T13:23:40.366396Z","title":"A framework for few-shot language model evaluation","venue":null,"work_id":"0cb338fe-f7b1-4751-8294-45b4a09056e0","year":2021},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.571833Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:c5c43c5e802733c8c0d4dea8e6898895588ec069b1326474ab65dc1f198be995","observation_id":"61a3804f-0fcc-4b29-918b-3fcd92a26dcb","resolution":{"observed_at":"2026-08-07T13:23:40.478669Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1609.07843","last_updated":"2016-09-26T04:06:13Z","snapshot_observed_at":"2026-07-06T05:12:10.387914Z","submitted_at":"2016-09-26T04:06:13Z","title":"Pointer Sentinel Mixture Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1609.07843","snapshot_observed_at":"2026-08-07T13:23:37.708800Z","title":"Pointer sentinel mixture models","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.708800Z"},"links":{"cited_paper":"/paper/1609.07843","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:7ccb9239a6e014ed580714625f2d9c1d929ad519ba85f3758f7c853b2befd8d1","observation_id":"1c39a6a7-394c-4d6e-b3d5-1f2faf654eb5","resolution":{"observed_at":"2026-08-07T13:23:37.708800Z","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":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-07T13:23:37.824977Z","title":"Exploring the limits of transfer learning with a unified text-to-text transformer","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.824977Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:13d12694a50ca67dd9b36fc874d50ab3f83d591b2fa48d2ff0700d822f36963d","observation_id":"26016e0a-1dd0-4cf4-beb2-c72ce6dded24","resolution":{"observed_at":"2026-08-07T13:23:37.824977Z","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":"2026-08-07T13:23:40.201994Z","title":"Boolq: Exploring the surprising difficulty of natural yes/no questions","venue":null,"work_id":"4cf8ce8b-e5a7-41c7-97c6-db1d988aaf62","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:37.966803Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:cc08a0ed450fc81212e288729be968f89a41977b76eb2ff1ebfb55b9c5a4c606","observation_id":"bce9becb-22dc-46f4-81e9-174271717929","resolution":{"observed_at":"2026-08-07T13:23:40.261827Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:39.969768Z","title":"Glue: A multi-task benchmark and analysis platform for natural language understanding","venue":null,"work_id":"90484339-36cb-427c-9761-7567ecfede86","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:38.092709Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:3b4f36820785e6f59b3efde184ab216bbc7a9e9e017ca42220e5e09e37d89d96","observation_id":"d97c6195-ea9d-453d-9768-533af0642a60","resolution":{"observed_at":"2026-08-07T13:23:40.082933Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:39.722883Z","title":"Hellaswag: Can a machine really finish your sentence? In ACL, 2019","venue":null,"work_id":"48faf916-1a63-45f4-85f0-32bf3197c3d1","year":2019},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:38.225259Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:329f8ef081dd31f970044a243c91cf80924362ccb0ad6354e976480399b14efb","observation_id":"5a9aee63-7109-4cc3-bb34-b27fa1fff767","resolution":{"observed_at":"2026-08-07T13:23:39.851000Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-07T13:23:39.517986Z","title":"Winogrande: An adversarial winograd schema challenge at scale","venue":null,"work_id":"89294fa9-cb2d-4924-9a68-6cd307a57aa5","year":2020},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:38.357206Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:c8d17c9c1061dbc60c8c5db1cab0c944365127139269cfee23ad0d7205610314","observation_id":"b8945154-e09f-4d53-a70a-6ad926a4365f","resolution":{"observed_at":"2026-08-07T13:23:39.599842Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1803.05457","last_updated":"2018-03-14T18:04:21Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2018-03-14T18:04:21Z","title":"Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1803.05457","snapshot_observed_at":"2026-08-07T13:23:38.469301Z","title":"Think you have solved question answering? try arc, the ai2 reasoning challenge","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:38.469301Z"},"links":{"cited_paper":"/paper/1803.05457","citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:a8d26b947741fbea748aeef77dad4697d1e17c0484e8647a51c88d45ce25b2fd","observation_id":"8514bbfd-9fa4-4620-91bb-107fe764d878","resolution":{"observed_at":"2026-08-07T13:23:38.469301Z","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":"2026-08-07T13:23:39.298765Z","title":"Can a suit of armor conduct electricity? a new dataset for open book question answering","venue":null,"work_id":"6618397e-4652-428f-bf10-e824dc68c22a","year":2018},"citing_paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-07T13:23:38.545260Z"},"links":{"citing_paper":"/paper/2505.21987"},"observation_digest":"sha256:6428bf5abc37d7eef64657a59198337dd047e86a13ef8509378f12d880995ea0","observation_id":"7e3dd250-68aa-4ad5-b4e5-cae178876883","resolution":{"observed_at":"2026-08-07T13:23:39.383951Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2505.21987","last_updated":"2025-05-28T05:25:16Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-09T09:35:04.687777Z","submitted_at":"2025-05-28T05:25:16Z","title":"ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning"},"reference_resolution":{"displayed":57,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":22,"verified_exact":2,"verified_fuzzy":33},"total_outbound_references":57},"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":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"thesis":"As of 9 August 2026, this Paper Citation Record lists 57 of 57 outbound references and 1 inbound Pith citation observation for arXiv:2505.21987."}