{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:X3P4TKVY47LSHIVUDRGBMVX36C","short_pith_number":"pith:X3P4TKVY","schema_version":"1.0","canonical_sha256":"bedfc9aab8e7d723a2b41c4c1656fbf0ba3210c2d363e05c4fcb1fef08eaac9f","source":{"kind":"arxiv","id":"2606.18961","version":1},"attestation_state":"computed","paper":{"title":"Be Your Own Teacher: Steering Protein Language Models via Unsupervised Reward Optimization","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lanqing Li, Pheng-Ann Heng, Shentong Mo, Yang Yu","submitted_at":"2026-06-17T11:42:01Z","abstract_excerpt":"Protein language models (PLMs) have emerged as powerful tools for controllable biomolecular design, yet their post-training adaptation typically relies on costly wet-lab validation or curated preference datasets. To overcome this supervision bottleneck, we introduce unsupervised reward optimization of PLMs, a comprehensive framework for steerable protein generation without ground-truth labels. Our key insight is that task-agnostic rewards, which combine intrinsic model uncertainty with extrinsic semantic consistency informed by protein representation models, exhibit strong correlation with con"},"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":"2606.18961","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-17T11:42:01Z","cross_cats_sorted":[],"title_canon_sha256":"7eb0c1f51e3c9c0f08782435c84d6e184259f25a95b7779f3ddddcf37b685332","abstract_canon_sha256":"c761da57a050be214bd517c300152d44ba93d45970b6eccf10aff6d0dc4628ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:11:53.050891Z","signature_b64":"2AXFA1rGXt0y78/h0RqC4dg46E7AWR0DQ5PSCj48b4K+vg6AEV/+zHBi+tAaslNVY9lM6P09Lj90Sn3WR/E5Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bedfc9aab8e7d723a2b41c4c1656fbf0ba3210c2d363e05c4fcb1fef08eaac9f","last_reissued_at":"2026-06-19T16:11:53.050530Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:11:53.050530Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Be Your Own Teacher: Steering Protein Language Models via Unsupervised Reward Optimization","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lanqing Li, Pheng-Ann Heng, Shentong Mo, Yang Yu","submitted_at":"2026-06-17T11:42:01Z","abstract_excerpt":"Protein language models (PLMs) have emerged as powerful tools for controllable biomolecular design, yet their post-training adaptation typically relies on costly wet-lab validation or curated preference datasets. To overcome this supervision bottleneck, we introduce unsupervised reward optimization of PLMs, a comprehensive framework for steerable protein generation without ground-truth labels. Our key insight is that task-agnostic rewards, which combine intrinsic model uncertainty with extrinsic semantic consistency informed by protein representation models, exhibit strong correlation with con"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.18961","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/2606.18961/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":"2606.18961","created_at":"2026-06-19T16:11:53.050591+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.18961v1","created_at":"2026-06-19T16:11:53.050591+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.18961","created_at":"2026-06-19T16:11:53.050591+00:00"},{"alias_kind":"pith_short_12","alias_value":"X3P4TKVY47LS","created_at":"2026-06-19T16:11:53.050591+00:00"},{"alias_kind":"pith_short_16","alias_value":"X3P4TKVY47LSHIVU","created_at":"2026-06-19T16:11:53.050591+00:00"},{"alias_kind":"pith_short_8","alias_value":"X3P4TKVY","created_at":"2026-06-19T16:11:53.050591+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/X3P4TKVY47LSHIVUDRGBMVX36C","json":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C.json","graph_json":"https://pith.science/api/pith-number/X3P4TKVY47LSHIVUDRGBMVX36C/graph.json","events_json":"https://pith.science/api/pith-number/X3P4TKVY47LSHIVUDRGBMVX36C/events.json","paper":"https://pith.science/paper/X3P4TKVY"},"agent_actions":{"view_html":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C","download_json":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C.json","view_paper":"https://pith.science/paper/X3P4TKVY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.18961&json=true","fetch_graph":"https://pith.science/api/pith-number/X3P4TKVY47LSHIVUDRGBMVX36C/graph.json","fetch_events":"https://pith.science/api/pith-number/X3P4TKVY47LSHIVUDRGBMVX36C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C/action/storage_attestation","attest_author":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C/action/author_attestation","sign_citation":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C/action/citation_signature","submit_replication":"https://pith.science/pith/X3P4TKVY47LSHIVUDRGBMVX36C/action/replication_record"}},"created_at":"2026-06-19T16:11:53.050591+00:00","updated_at":"2026-06-19T16:11:53.050591+00:00"}