{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KAFUXR4A2I2D4VVN2GDMZHQJNW","short_pith_number":"pith:KAFUXR4A","schema_version":"1.0","canonical_sha256":"500b4bc780d2343e56add186cc9e096d95084e8abbce7c182ac7501f1ecedbfd","source":{"kind":"arxiv","id":"2506.10341","version":2},"attestation_state":"computed","paper":{"title":"Formalizing Learning from Language Feedback with Provable Guarantees","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Adith Swaminathan, Aditya Modi, Allen Nie, Ching-An Cheng, Ruijie Zheng, Wanqiao Xu","submitted_at":"2025-06-12T04:35:02Z","abstract_excerpt":"Interactively learning from observation and language feedback is an increasingly studied area driven by the emergence of large language model (LLM) agents. Despite impressive empirical demonstrations, so far a principled framing of these decision problems remains lacking. We formalize the Learning from Language Feedback (LLF) problem, assert sufficient assumptions to enable learning despite latent rewards, and introduce $\\textit{transfer eluder dimension}$ as a measure to characterize the hardness of LLF. We formalize the intuition that information in the language feedback governs the learning"},"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":"2506.10341","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-12T04:35:02Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"2d35fed922642af0a3900df2a1d4174ccba488d94239df71f38df40e9cc37036","abstract_canon_sha256":"4107e26d678b39c880178c4abfb9816d052787481783ae1c8bba226d49c8574c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:07.149285Z","signature_b64":"BrNcIsenUUC1pfL4dVlwJt6DIslCM2cQGV9a9zmFCQD76/rQPh3zycMNWD3grgethMph3m/vb64Eu2y+mYSvBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"500b4bc780d2343e56add186cc9e096d95084e8abbce7c182ac7501f1ecedbfd","last_reissued_at":"2026-06-09T01:05:07.148466Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:07.148466Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Formalizing Learning from Language Feedback with Provable Guarantees","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Adith Swaminathan, Aditya Modi, Allen Nie, Ching-An Cheng, Ruijie Zheng, Wanqiao Xu","submitted_at":"2025-06-12T04:35:02Z","abstract_excerpt":"Interactively learning from observation and language feedback is an increasingly studied area driven by the emergence of large language model (LLM) agents. Despite impressive empirical demonstrations, so far a principled framing of these decision problems remains lacking. We formalize the Learning from Language Feedback (LLF) problem, assert sufficient assumptions to enable learning despite latent rewards, and introduce $\\textit{transfer eluder dimension}$ as a measure to characterize the hardness of LLF. We formalize the intuition that information in the language feedback governs the learning"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.10341","kind":"arxiv","version":2},"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/2506.10341/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":"2506.10341","created_at":"2026-06-09T01:05:07.148698+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.10341v2","created_at":"2026-06-09T01:05:07.148698+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.10341","created_at":"2026-06-09T01:05:07.148698+00:00"},{"alias_kind":"pith_short_12","alias_value":"KAFUXR4A2I2D","created_at":"2026-06-09T01:05:07.148698+00:00"},{"alias_kind":"pith_short_16","alias_value":"KAFUXR4A2I2D4VVN","created_at":"2026-06-09T01:05:07.148698+00:00"},{"alias_kind":"pith_short_8","alias_value":"KAFUXR4A","created_at":"2026-06-09T01:05:07.148698+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.15113","citing_title":"Learning from Language Feedback via Variational Policy Distillation","ref_index":37,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW","json":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW.json","graph_json":"https://pith.science/api/pith-number/KAFUXR4A2I2D4VVN2GDMZHQJNW/graph.json","events_json":"https://pith.science/api/pith-number/KAFUXR4A2I2D4VVN2GDMZHQJNW/events.json","paper":"https://pith.science/paper/KAFUXR4A"},"agent_actions":{"view_html":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW","download_json":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW.json","view_paper":"https://pith.science/paper/KAFUXR4A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.10341&json=true","fetch_graph":"https://pith.science/api/pith-number/KAFUXR4A2I2D4VVN2GDMZHQJNW/graph.json","fetch_events":"https://pith.science/api/pith-number/KAFUXR4A2I2D4VVN2GDMZHQJNW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW/action/storage_attestation","attest_author":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW/action/author_attestation","sign_citation":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW/action/citation_signature","submit_replication":"https://pith.science/pith/KAFUXR4A2I2D4VVN2GDMZHQJNW/action/replication_record"}},"created_at":"2026-06-09T01:05:07.148698+00:00","updated_at":"2026-06-09T01:05:07.148698+00:00"}