{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:E55TT6MIPEPGMPXA6BMHLKFMIU","short_pith_number":"pith:E55TT6MI","schema_version":"1.0","canonical_sha256":"277b39f988791e663ee0f05875a8ac453b8a403189bd7875e075b98e0ea2d3f8","source":{"kind":"arxiv","id":"2606.12609","version":1},"attestation_state":"computed","paper":{"title":"Viral Proteins Reveal Geometry of Protein Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["q-bio.QM"],"primary_cat":"cs.LG","authors_text":"Arthur Bigot, Core Francisco Park, Dianzhuo Wang, Eugene Shakhnovich, Harmon Bhasin","submitted_at":"2026-06-10T19:04:34Z","abstract_excerpt":"Protein language models are trained on highly imbalanced datasets, raising the question of how they represent underrepresented biological sequences. Using viral proteins as a case study across ESM model families, we identify a dominant nativeness axis in embedding space, aligned with masked reconstruction perplexity, that orders sequences from well-modeled cellular proteins through viral proteins to shuffled and random sequences. Scaling contracts this axis unevenly across viral families. Despite this, protein language model embeddings retain viral-specific signal: viral proteins remain linear"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.12609","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-10T19:04:34Z","cross_cats_sorted":["q-bio.QM"],"title_canon_sha256":"ae4dd9afaa958182f5ab538f8a06900b9b6e0c2a0df091d5b8fb17a2a359f737","abstract_canon_sha256":"28c7ad61d9d634dd9040dc11cf5c82b9ef342c9e9c9e7898d36965c38af7c4dd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T00:07:56.290550Z","signature_b64":"kOmtsp3im4fWZJ59GnqDpCohqM6DUKcUFDWJDrHGIRBL0rtrbhmeGX1jnamTbK8cHaThC3tzhwONV+VCTpAYDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"277b39f988791e663ee0f05875a8ac453b8a403189bd7875e075b98e0ea2d3f8","last_reissued_at":"2026-06-12T00:07:56.290148Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T00:07:56.290148Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Viral Proteins Reveal Geometry of Protein Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["q-bio.QM"],"primary_cat":"cs.LG","authors_text":"Arthur Bigot, Core Francisco Park, Dianzhuo Wang, Eugene Shakhnovich, Harmon Bhasin","submitted_at":"2026-06-10T19:04:34Z","abstract_excerpt":"Protein language models are trained on highly imbalanced datasets, raising the question of how they represent underrepresented biological sequences. Using viral proteins as a case study across ESM model families, we identify a dominant nativeness axis in embedding space, aligned with masked reconstruction perplexity, that orders sequences from well-modeled cellular proteins through viral proteins to shuffled and random sequences. Scaling contracts this axis unevenly across viral families. Despite this, protein language model embeddings retain viral-specific signal: viral proteins remain linear"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12609","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.12609/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.12609","created_at":"2026-06-12T00:07:56.290205+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.12609v1","created_at":"2026-06-12T00:07:56.290205+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12609","created_at":"2026-06-12T00:07:56.290205+00:00"},{"alias_kind":"pith_short_12","alias_value":"E55TT6MIPEPG","created_at":"2026-06-12T00:07:56.290205+00:00"},{"alias_kind":"pith_short_16","alias_value":"E55TT6MIPEPGMPXA","created_at":"2026-06-12T00:07:56.290205+00:00"},{"alias_kind":"pith_short_8","alias_value":"E55TT6MI","created_at":"2026-06-12T00:07:56.290205+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU","json":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU.json","graph_json":"https://pith.science/api/pith-number/E55TT6MIPEPGMPXA6BMHLKFMIU/graph.json","events_json":"https://pith.science/api/pith-number/E55TT6MIPEPGMPXA6BMHLKFMIU/events.json","paper":"https://pith.science/paper/E55TT6MI"},"agent_actions":{"view_html":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU","download_json":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU.json","view_paper":"https://pith.science/paper/E55TT6MI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.12609&json=true","fetch_graph":"https://pith.science/api/pith-number/E55TT6MIPEPGMPXA6BMHLKFMIU/graph.json","fetch_events":"https://pith.science/api/pith-number/E55TT6MIPEPGMPXA6BMHLKFMIU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU/action/storage_attestation","attest_author":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU/action/author_attestation","sign_citation":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU/action/citation_signature","submit_replication":"https://pith.science/pith/E55TT6MIPEPGMPXA6BMHLKFMIU/action/replication_record"}},"created_at":"2026-06-12T00:07:56.290205+00:00","updated_at":"2026-06-12T00:07:56.290205+00:00"}