{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:JPSVGONLMIUEFEPCAQKOAZ76EZ","short_pith_number":"pith:JPSVGONL","schema_version":"1.0","canonical_sha256":"4be55339ab62284291e20414e067fe267eb9528410b11137f893810806ea25a8","source":{"kind":"arxiv","id":"2305.10314","version":2},"attestation_state":"computed","paper":{"title":"LeTI: Learning to Generate from Textual Interactions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.CL","authors_text":"Hao Peng, Heng Ji, Reyhaneh Jabbarvand, Xingyao Wang","submitted_at":"2023-05-17T15:53:31Z","abstract_excerpt":"Fine-tuning pre-trained language models (LMs) is essential for enhancing their capabilities. Existing techniques commonly fine-tune on input-output pairs (e.g., instruction tuning) or with numerical rewards that gauge the output quality (e.g., RLHF). We explore LMs' potential to learn from textual interactions (LETI) that not only check their correctness with binary labels but also pinpoint and explain errors in their outputs through textual feedback. Our focus is the code generation task, where the model produces code based on natural language instructions. This setting invites a natural and "},"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":"2305.10314","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-17T15:53:31Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"2a871d937a4d6aad3fdbb8ee8c8f72ce72a91dd3f3b2f70b0d63b1c7b5925743","abstract_canon_sha256":"4a9194e173c28da0976b884ef0d63b0697e3e2502b7cafbdbf1f5f4bc0005b54"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:57:40.490492Z","signature_b64":"rUikM05j47g4Yan+9IlEcAop6YxqESix39O/5CuBFjduY+rCxQ9Jeh9kdvQKtHaWD2fMhyIn8ILUzI8IYAwVAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4be55339ab62284291e20414e067fe267eb9528410b11137f893810806ea25a8","last_reissued_at":"2026-07-05T07:57:40.489981Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:57:40.489981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LeTI: Learning to Generate from Textual Interactions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.CL","authors_text":"Hao Peng, Heng Ji, Reyhaneh Jabbarvand, Xingyao Wang","submitted_at":"2023-05-17T15:53:31Z","abstract_excerpt":"Fine-tuning pre-trained language models (LMs) is essential for enhancing their capabilities. Existing techniques commonly fine-tune on input-output pairs (e.g., instruction tuning) or with numerical rewards that gauge the output quality (e.g., RLHF). We explore LMs' potential to learn from textual interactions (LETI) that not only check their correctness with binary labels but also pinpoint and explain errors in their outputs through textual feedback. Our focus is the code generation task, where the model produces code based on natural language instructions. This setting invites a natural and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.10314","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/2305.10314/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":"2305.10314","created_at":"2026-07-05T07:57:40.490040+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.10314v2","created_at":"2026-07-05T07:57:40.490040+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.10314","created_at":"2026-07-05T07:57:40.490040+00:00"},{"alias_kind":"pith_short_12","alias_value":"JPSVGONLMIUE","created_at":"2026-07-05T07:57:40.490040+00:00"},{"alias_kind":"pith_short_16","alias_value":"JPSVGONLMIUEFEPC","created_at":"2026-07-05T07:57:40.490040+00:00"},{"alias_kind":"pith_short_8","alias_value":"JPSVGONL","created_at":"2026-07-05T07:57:40.490040+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.07599","citing_title":"Stencil Computations on Tenstorrent Wormhole","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ","json":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ.json","graph_json":"https://pith.science/api/pith-number/JPSVGONLMIUEFEPCAQKOAZ76EZ/graph.json","events_json":"https://pith.science/api/pith-number/JPSVGONLMIUEFEPCAQKOAZ76EZ/events.json","paper":"https://pith.science/paper/JPSVGONL"},"agent_actions":{"view_html":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ","download_json":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ.json","view_paper":"https://pith.science/paper/JPSVGONL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.10314&json=true","fetch_graph":"https://pith.science/api/pith-number/JPSVGONLMIUEFEPCAQKOAZ76EZ/graph.json","fetch_events":"https://pith.science/api/pith-number/JPSVGONLMIUEFEPCAQKOAZ76EZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ/action/storage_attestation","attest_author":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ/action/author_attestation","sign_citation":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ/action/citation_signature","submit_replication":"https://pith.science/pith/JPSVGONLMIUEFEPCAQKOAZ76EZ/action/replication_record"}},"created_at":"2026-07-05T07:57:40.490040+00:00","updated_at":"2026-07-05T07:57:40.490040+00:00"}