{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:LAEOYYRHVT6YLKPDAXGQR2G7FA","short_pith_number":"pith:LAEOYYRH","canonical_record":{"source":{"id":"2605.14893","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T14:37:50Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"6dede2f5090499c5eccea9195b931d00ebfc92e1a3f9902155e3bdba2fecc229","abstract_canon_sha256":"769945c003f1900799ce46ec338bc71ba7a855a14ecb33e8728244be4c471690"},"schema_version":"1.0"},"canonical_sha256":"5808ec6227acfd85a9e305cd08e8df2804827d843f836d9cec1d1b1aeff73193","source":{"kind":"arxiv","id":"2605.14893","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14893","created_at":"2026-05-17T23:38:55Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14893v1","created_at":"2026-05-17T23:38:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14893","created_at":"2026-05-17T23:38:55Z"},{"alias_kind":"pith_short_12","alias_value":"LAEOYYRHVT6Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LAEOYYRHVT6YLKPD","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LAEOYYRH","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:LAEOYYRHVT6YLKPDAXGQR2G7FA","target":"record","payload":{"canonical_record":{"source":{"id":"2605.14893","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T14:37:50Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"6dede2f5090499c5eccea9195b931d00ebfc92e1a3f9902155e3bdba2fecc229","abstract_canon_sha256":"769945c003f1900799ce46ec338bc71ba7a855a14ecb33e8728244be4c471690"},"schema_version":"1.0"},"canonical_sha256":"5808ec6227acfd85a9e305cd08e8df2804827d843f836d9cec1d1b1aeff73193","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:55.920498Z","signature_b64":"icPicJ1t+oh1U/UXd1CI4YthCZNmk5vq7AyAub0ab8EgmwvbkW8Z27GjyKO6DNJXBNVtgUgVrxsdbR3Mr3aaCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5808ec6227acfd85a9e305cd08e8df2804827d843f836d9cec1d1b1aeff73193","last_reissued_at":"2026-05-17T23:38:55.919824Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:55.919824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.14893","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:38:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tvCfCtHgB+pbkq8H6FKmAMaBgsgK2Fx0UeW1cnfBzo9my/jFgmP55Fuq5cWWaTLyGLppFaSITBhMDttXKafwAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T23:09:43.914648Z"},"content_sha256":"c3cfa708fa9492da0234cfde33c98dd3c0fa66902290fc6e1b3a5c5fce3649aa","schema_version":"1.0","event_id":"sha256:c3cfa708fa9492da0234cfde33c98dd3c0fa66902290fc6e1b3a5c5fce3649aa"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:LAEOYYRHVT6YLKPDAXGQR2G7FA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Your CLIP has 164 dimensions of noise: Exploring the embeddings covariance eigenspectrum of contrastively pretrained vision-language transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Dawid P{\\l}udowski, Jakub Grzywaczewski, Przemys{\\l}aw Biecek","submitted_at":"2026-05-14T14:37:50Z","abstract_excerpt":"Contrastively pre-trained Vision-Language Models (VLMs) serve as powerful feature extractors. Yet, their shared latent spaces are prone to structural anomalies and act as repositories for non-semantic, multi-modal noise. To address this phenomenon, we employ spectral decomposition of covariance matrices to decompose the VLM latent space into a multi-modal semantic signal component and a shared noise subspace. We observe that this noise geometry exhibits strong subgroup invariance across distinct data subsets. Crucially, pruning these shared noise dimensions is mainly harmless, preserving or ac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.14893","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:38:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CCDNAQTaAWLEe7TIRCDE17+oz7JSHhBnJtLXzoAyfpKZ4ZhdtPRFiIVcbfrgBTBGVP42gp7hCZJ9ppvm9uKQBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T23:09:43.914987Z"},"content_sha256":"876ffa2162406ef1ba293d97fb514c7e76e600bfab83a0317fa93b527a003fb1","schema_version":"1.0","event_id":"sha256:876ffa2162406ef1ba293d97fb514c7e76e600bfab83a0317fa93b527a003fb1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA/bundle.json","state_url":"https://pith.science/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-29T23:09:43Z","links":{"resolver":"https://pith.science/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA","bundle":"https://pith.science/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA/bundle.json","state":"https://pith.science/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LAEOYYRHVT6YLKPDAXGQR2G7FA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:LAEOYYRHVT6YLKPDAXGQR2G7FA","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"769945c003f1900799ce46ec338bc71ba7a855a14ecb33e8728244be4c471690","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T14:37:50Z","title_canon_sha256":"6dede2f5090499c5eccea9195b931d00ebfc92e1a3f9902155e3bdba2fecc229"},"schema_version":"1.0","source":{"id":"2605.14893","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14893","created_at":"2026-05-17T23:38:55Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14893v1","created_at":"2026-05-17T23:38:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14893","created_at":"2026-05-17T23:38:55Z"},{"alias_kind":"pith_short_12","alias_value":"LAEOYYRHVT6Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LAEOYYRHVT6YLKPD","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LAEOYYRH","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:876ffa2162406ef1ba293d97fb514c7e76e600bfab83a0317fa93b527a003fb1","target":"graph","created_at":"2026-05-17T23:38:55Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Contrastively pre-trained Vision-Language Models (VLMs) serve as powerful feature extractors. Yet, their shared latent spaces are prone to structural anomalies and act as repositories for non-semantic, multi-modal noise. To address this phenomenon, we employ spectral decomposition of covariance matrices to decompose the VLM latent space into a multi-modal semantic signal component and a shared noise subspace. We observe that this noise geometry exhibits strong subgroup invariance across distinct data subsets. Crucially, pruning these shared noise dimensions is mainly harmless, preserving or ac","authors_text":"Dawid P{\\l}udowski, Jakub Grzywaczewski, Przemys{\\l}aw Biecek","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T14:37:50Z","title":"Your CLIP has 164 dimensions of noise: Exploring the embeddings covariance eigenspectrum of contrastively pretrained vision-language transformers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.14893","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:c3cfa708fa9492da0234cfde33c98dd3c0fa66902290fc6e1b3a5c5fce3649aa","target":"record","created_at":"2026-05-17T23:38:55Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"769945c003f1900799ce46ec338bc71ba7a855a14ecb33e8728244be4c471690","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T14:37:50Z","title_canon_sha256":"6dede2f5090499c5eccea9195b931d00ebfc92e1a3f9902155e3bdba2fecc229"},"schema_version":"1.0","source":{"id":"2605.14893","kind":"arxiv","version":1}},"canonical_sha256":"5808ec6227acfd85a9e305cd08e8df2804827d843f836d9cec1d1b1aeff73193","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5808ec6227acfd85a9e305cd08e8df2804827d843f836d9cec1d1b1aeff73193","first_computed_at":"2026-05-17T23:38:55.919824Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:55.919824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"icPicJ1t+oh1U/UXd1CI4YthCZNmk5vq7AyAub0ab8EgmwvbkW8Z27GjyKO6DNJXBNVtgUgVrxsdbR3Mr3aaCg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:55.920498Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.14893","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c3cfa708fa9492da0234cfde33c98dd3c0fa66902290fc6e1b3a5c5fce3649aa","sha256:876ffa2162406ef1ba293d97fb514c7e76e600bfab83a0317fa93b527a003fb1"],"state_sha256":"4c5fe81b753d25daaecc75c563a20ee0f0b85f16df3ec8dbfa99e6ed777fe6dc"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zQwzJglekNsX8Jhs6PWmPlxVXzcQuMX8rxtCDzbTw7lBMugv2PN75CUFK4GZtxa7JI3BXuVx6QkFiVZKmcYjCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T23:09:43.917578Z","bundle_sha256":"50be363b2567eca4db70ab086eb71ba35313a8b85e62fad719b314d1c261681f"}}