{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:4ERCPANKWGDOYFTDIHBCQBDQWS","short_pith_number":"pith:4ERCPANK","canonical_record":{"source":{"id":"2412.12101","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"q-bio.BM","submitted_at":"2024-11-13T18:51:21Z","cross_cats_sorted":["cs.AI","cs.LG","q-bio.QM"],"title_canon_sha256":"afe7f8484c323d15ad50d6aa943939bae4134a3863b2cc2c1315dc49e93d7bfa","abstract_canon_sha256":"80bada6eeda4c5a1aa756ad955111754d5ed4fdf0f9dc3340c3b16b9a7e5e25a"},"schema_version":"1.0"},"canonical_sha256":"e1222781aab186ec166341c2280470b4915dfc1c7e84a503a1c3aa316432994e","source":{"kind":"arxiv","id":"2412.12101","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.12101","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"arxiv_version","alias_value":"2412.12101v1","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.12101","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"pith_short_12","alias_value":"4ERCPANKWGDO","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"pith_short_16","alias_value":"4ERCPANKWGDOYFTD","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"pith_short_8","alias_value":"4ERCPANK","created_at":"2026-07-05T09:50:09Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:4ERCPANKWGDOYFTDIHBCQBDQWS","target":"record","payload":{"canonical_record":{"source":{"id":"2412.12101","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"q-bio.BM","submitted_at":"2024-11-13T18:51:21Z","cross_cats_sorted":["cs.AI","cs.LG","q-bio.QM"],"title_canon_sha256":"afe7f8484c323d15ad50d6aa943939bae4134a3863b2cc2c1315dc49e93d7bfa","abstract_canon_sha256":"80bada6eeda4c5a1aa756ad955111754d5ed4fdf0f9dc3340c3b16b9a7e5e25a"},"schema_version":"1.0"},"canonical_sha256":"e1222781aab186ec166341c2280470b4915dfc1c7e84a503a1c3aa316432994e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:50:09.688722Z","signature_b64":"hmZxTwiWEsXhx0S94h2YTLsjw6gJddgf7LbhPZS6artPbZr0E2m/r06vVDwvbA50bh4m9V5NYkrDSNmFe529BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e1222781aab186ec166341c2280470b4915dfc1c7e84a503a1c3aa316432994e","last_reissued_at":"2026-07-05T09:50:09.688252Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:50:09.688252Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.12101","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":"2026-07-05T09:50:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7QvNak9XFhuJ2BIUNvfJ4ZKelMAEm1/hwwuFuVB5CT/uR6eO4ot/kRkfLoYtW3uKbKgb9GgS0zEOvJKs+F6JDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-05T14:13:39.123734Z"},"content_sha256":"e79741a0e18fa3e733b23d62049f7a6ebad151474c882d07e41ebc50bc837ae2","schema_version":"1.0","event_id":"sha256:e79741a0e18fa3e733b23d62049f7a6ebad151474c882d07e41ebc50bc837ae2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:4ERCPANKWGDOYFTDIHBCQBDQWS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","q-bio.QM"],"primary_cat":"q-bio.BM","authors_text":"Elana Simon, James Zou","submitted_at":"2024-11-13T18:51:21Z","abstract_excerpt":"Protein language models (PLMs) have demonstrated remarkable success in protein modeling and design, yet their internal mechanisms for predicting structure and function remain poorly understood. Here we present a systematic approach to extract and analyze interpretable features from PLMs using sparse autoencoders (SAEs). By training SAEs on embeddings from the PLM ESM-2, we identify up to 2,548 human-interpretable latent features per layer that strongly correlate with up to 143 known biological concepts such as binding sites, structural motifs, and functional domains. In contrast, examining ind"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.12101","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/2412.12101/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:50:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"plp/7+qqI9HPljK76AaCuv3JLDsKNkqrdhjl7VW4lzc4eblXkYafgBCYmoBgs6+Jd3DVvLNpkeUsctyLwHgSDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-05T14:13:39.124397Z"},"content_sha256":"7247d02722ac8e6ae650d321f616ec0ef46fdf60d0c9f0a65bc8e4d1674ac2a7","schema_version":"1.0","event_id":"sha256:7247d02722ac8e6ae650d321f616ec0ef46fdf60d0c9f0a65bc8e4d1674ac2a7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4ERCPANKWGDOYFTDIHBCQBDQWS/bundle.json","state_url":"https://pith.science/pith/4ERCPANKWGDOYFTDIHBCQBDQWS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4ERCPANKWGDOYFTDIHBCQBDQWS/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":"2026-07-05T14:13:39Z","links":{"resolver":"https://pith.science/pith/4ERCPANKWGDOYFTDIHBCQBDQWS","bundle":"https://pith.science/pith/4ERCPANKWGDOYFTDIHBCQBDQWS/bundle.json","state":"https://pith.science/pith/4ERCPANKWGDOYFTDIHBCQBDQWS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4ERCPANKWGDOYFTDIHBCQBDQWS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:4ERCPANKWGDOYFTDIHBCQBDQWS","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":"80bada6eeda4c5a1aa756ad955111754d5ed4fdf0f9dc3340c3b16b9a7e5e25a","cross_cats_sorted":["cs.AI","cs.LG","q-bio.QM"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"q-bio.BM","submitted_at":"2024-11-13T18:51:21Z","title_canon_sha256":"afe7f8484c323d15ad50d6aa943939bae4134a3863b2cc2c1315dc49e93d7bfa"},"schema_version":"1.0","source":{"id":"2412.12101","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.12101","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"arxiv_version","alias_value":"2412.12101v1","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.12101","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"pith_short_12","alias_value":"4ERCPANKWGDO","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"pith_short_16","alias_value":"4ERCPANKWGDOYFTD","created_at":"2026-07-05T09:50:09Z"},{"alias_kind":"pith_short_8","alias_value":"4ERCPANK","created_at":"2026-07-05T09:50:09Z"}],"graph_snapshots":[{"event_id":"sha256:7247d02722ac8e6ae650d321f616ec0ef46fdf60d0c9f0a65bc8e4d1674ac2a7","target":"graph","created_at":"2026-07-05T09:50:09Z","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"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2412.12101/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Protein language models (PLMs) have demonstrated remarkable success in protein modeling and design, yet their internal mechanisms for predicting structure and function remain poorly understood. Here we present a systematic approach to extract and analyze interpretable features from PLMs using sparse autoencoders (SAEs). By training SAEs on embeddings from the PLM ESM-2, we identify up to 2,548 human-interpretable latent features per layer that strongly correlate with up to 143 known biological concepts such as binding sites, structural motifs, and functional domains. In contrast, examining ind","authors_text":"Elana Simon, James Zou","cross_cats":["cs.AI","cs.LG","q-bio.QM"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"q-bio.BM","submitted_at":"2024-11-13T18:51:21Z","title":"InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.12101","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:e79741a0e18fa3e733b23d62049f7a6ebad151474c882d07e41ebc50bc837ae2","target":"record","created_at":"2026-07-05T09:50:09Z","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":"80bada6eeda4c5a1aa756ad955111754d5ed4fdf0f9dc3340c3b16b9a7e5e25a","cross_cats_sorted":["cs.AI","cs.LG","q-bio.QM"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"q-bio.BM","submitted_at":"2024-11-13T18:51:21Z","title_canon_sha256":"afe7f8484c323d15ad50d6aa943939bae4134a3863b2cc2c1315dc49e93d7bfa"},"schema_version":"1.0","source":{"id":"2412.12101","kind":"arxiv","version":1}},"canonical_sha256":"e1222781aab186ec166341c2280470b4915dfc1c7e84a503a1c3aa316432994e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e1222781aab186ec166341c2280470b4915dfc1c7e84a503a1c3aa316432994e","first_computed_at":"2026-07-05T09:50:09.688252Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:50:09.688252Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hmZxTwiWEsXhx0S94h2YTLsjw6gJddgf7LbhPZS6artPbZr0E2m/r06vVDwvbA50bh4m9V5NYkrDSNmFe529BA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:50:09.688722Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.12101","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e79741a0e18fa3e733b23d62049f7a6ebad151474c882d07e41ebc50bc837ae2","sha256:7247d02722ac8e6ae650d321f616ec0ef46fdf60d0c9f0a65bc8e4d1674ac2a7"],"state_sha256":"394da19f62c1aab53f3d6e2df8af5ef996afe0036bdedc5d8cc0c7bb5249aa67"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WxQhoKNvKLwkuPoBijP324B5K4CU0k+UnHBkMzd0Y3mJDjUUlRClwGTMnDUrQ+P0cQmbjRJR/2rjdRUIIlZeAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-05T14:13:39.128550Z","bundle_sha256":"10c7b2380838c3e3b7f2d3bd9fb14629372280966832719807db870081363de6"}}