{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:HEVUYQTAJTWMOMOTMCZMUK4OZR","short_pith_number":"pith:HEVUYQTA","schema_version":"1.0","canonical_sha256":"392b4c42604cecc731d360b2ca2b8ecc7223f98ead6ea8cefdfc7ee7ba0b1d08","source":{"kind":"arxiv","id":"2306.01150","version":1},"attestation_state":"computed","paper":{"title":"Did You Read the Instructions? Rethinking the Effectiveness of Task Definitions in Instruction Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Caiming Xiong, Chien-Sheng Jason Wu, Fan Yin, Jesse Vig, Philippe Laban, Shafiq Joty","submitted_at":"2023-06-01T21:11:24Z","abstract_excerpt":"Large language models (LLMs) have shown impressive performance in following natural language instructions to solve unseen tasks. However, it remains unclear whether models truly understand task definitions and whether the human-written definitions are optimal. In this paper, we systematically study the role of task definitions in instruction learning. We first conduct an ablation analysis informed by human annotations to understand which parts of a task definition are most important, and find that model performance only drops substantially when removing contents describing the task output, in "},"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":"2306.01150","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-01T21:11:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"154f931a9799d6e3887e833aab466f0d05e8169aea1ca9d09ff7ea37b5fe70e2","abstract_canon_sha256":"9ebba4ef9f62800c366e6cd22ecc0b96c5d6505a623b35802bbda531a7c06e6d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:16:38.698905Z","signature_b64":"09U3vYHnx937lCewARm0oxvNPqa29ofjQz63sA2w/z8ockLl7cN3rqZ/NPIfL/TqHLopaiRdx96d0pyAZIgFCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"392b4c42604cecc731d360b2ca2b8ecc7223f98ead6ea8cefdfc7ee7ba0b1d08","last_reissued_at":"2026-07-05T06:16:38.698495Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:16:38.698495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Did You Read the Instructions? Rethinking the Effectiveness of Task Definitions in Instruction Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Caiming Xiong, Chien-Sheng Jason Wu, Fan Yin, Jesse Vig, Philippe Laban, Shafiq Joty","submitted_at":"2023-06-01T21:11:24Z","abstract_excerpt":"Large language models (LLMs) have shown impressive performance in following natural language instructions to solve unseen tasks. However, it remains unclear whether models truly understand task definitions and whether the human-written definitions are optimal. In this paper, we systematically study the role of task definitions in instruction learning. We first conduct an ablation analysis informed by human annotations to understand which parts of a task definition are most important, and find that model performance only drops substantially when removing contents describing the task output, in "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.01150","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/2306.01150/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":"2306.01150","created_at":"2026-07-05T06:16:38.698551+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.01150v1","created_at":"2026-07-05T06:16:38.698551+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.01150","created_at":"2026-07-05T06:16:38.698551+00:00"},{"alias_kind":"pith_short_12","alias_value":"HEVUYQTAJTWM","created_at":"2026-07-05T06:16:38.698551+00:00"},{"alias_kind":"pith_short_16","alias_value":"HEVUYQTAJTWMOMOT","created_at":"2026-07-05T06:16:38.698551+00:00"},{"alias_kind":"pith_short_8","alias_value":"HEVUYQTA","created_at":"2026-07-05T06:16:38.698551+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2404.14294","citing_title":"A Survey on Efficient Inference for Large Language Models","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR","json":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR.json","graph_json":"https://pith.science/api/pith-number/HEVUYQTAJTWMOMOTMCZMUK4OZR/graph.json","events_json":"https://pith.science/api/pith-number/HEVUYQTAJTWMOMOTMCZMUK4OZR/events.json","paper":"https://pith.science/paper/HEVUYQTA"},"agent_actions":{"view_html":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR","download_json":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR.json","view_paper":"https://pith.science/paper/HEVUYQTA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.01150&json=true","fetch_graph":"https://pith.science/api/pith-number/HEVUYQTAJTWMOMOTMCZMUK4OZR/graph.json","fetch_events":"https://pith.science/api/pith-number/HEVUYQTAJTWMOMOTMCZMUK4OZR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR/action/storage_attestation","attest_author":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR/action/author_attestation","sign_citation":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR/action/citation_signature","submit_replication":"https://pith.science/pith/HEVUYQTAJTWMOMOTMCZMUK4OZR/action/replication_record"}},"created_at":"2026-07-05T06:16:38.698551+00:00","updated_at":"2026-07-05T06:16:38.698551+00:00"}