{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HR5W6PUJRZU6KPK7T5W2GOHUG5","short_pith_number":"pith:HR5W6PUJ","schema_version":"1.0","canonical_sha256":"3c7b6f3e898e69e53d5f9f6da338f4374321861807b8e8e5d0bb1e2490560d9a","source":{"kind":"arxiv","id":"2506.15629","version":1},"attestation_state":"computed","paper":{"title":"Revisiting Compositional Generalization Capability of Large Language Models Considering Instruction Following Ability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hidetaka Kamigaito, Taro Watanabe, Yusuke Sakai","submitted_at":"2025-06-18T17:00:54Z","abstract_excerpt":"In generative commonsense reasoning tasks such as CommonGen, generative large language models (LLMs) compose sentences that include all given concepts. However, when focusing on instruction-following capabilities, if a prompt specifies a concept order, LLMs must generate sentences that adhere to the specified order. To address this, we propose Ordered CommonGen, a benchmark designed to evaluate the compositional generalization and instruction-following abilities of LLMs. This benchmark measures ordered coverage to assess whether concepts are generated in the specified order, enabling a simulta"},"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":"2506.15629","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T17:00:54Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a1b969bbe5dd0b1e6c9d1f1d1273fa2d71e77f0e8e1fcaabb87c5756e8b03185","abstract_canon_sha256":"c6c24a81a32f76d1764719b10f5099d8f8160bf022a7e080200f81d1fbccf464"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:47.333433Z","signature_b64":"swWJ5Vjiti1mW/0m2Sr1M9sBYcA3iMeo0VSWv+aorIuck0t62VR05sWk9Izfa7oYRhtGW1mkHbZEwhoRqBNIAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3c7b6f3e898e69e53d5f9f6da338f4374321861807b8e8e5d0bb1e2490560d9a","last_reissued_at":"2026-07-05T11:23:47.332861Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:47.332861Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting Compositional Generalization Capability of Large Language Models Considering Instruction Following Ability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hidetaka Kamigaito, Taro Watanabe, Yusuke Sakai","submitted_at":"2025-06-18T17:00:54Z","abstract_excerpt":"In generative commonsense reasoning tasks such as CommonGen, generative large language models (LLMs) compose sentences that include all given concepts. However, when focusing on instruction-following capabilities, if a prompt specifies a concept order, LLMs must generate sentences that adhere to the specified order. To address this, we propose Ordered CommonGen, a benchmark designed to evaluate the compositional generalization and instruction-following abilities of LLMs. This benchmark measures ordered coverage to assess whether concepts are generated in the specified order, enabling a simulta"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15629","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/2506.15629/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":"2506.15629","created_at":"2026-07-05T11:23:47.332936+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.15629v1","created_at":"2026-07-05T11:23:47.332936+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15629","created_at":"2026-07-05T11:23:47.332936+00:00"},{"alias_kind":"pith_short_12","alias_value":"HR5W6PUJRZU6","created_at":"2026-07-05T11:23:47.332936+00:00"},{"alias_kind":"pith_short_16","alias_value":"HR5W6PUJRZU6KPK7","created_at":"2026-07-05T11:23:47.332936+00:00"},{"alias_kind":"pith_short_8","alias_value":"HR5W6PUJ","created_at":"2026-07-05T11:23:47.332936+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.05139","citing_title":"Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning","ref_index":42,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5","json":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5.json","graph_json":"https://pith.science/api/pith-number/HR5W6PUJRZU6KPK7T5W2GOHUG5/graph.json","events_json":"https://pith.science/api/pith-number/HR5W6PUJRZU6KPK7T5W2GOHUG5/events.json","paper":"https://pith.science/paper/HR5W6PUJ"},"agent_actions":{"view_html":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5","download_json":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5.json","view_paper":"https://pith.science/paper/HR5W6PUJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.15629&json=true","fetch_graph":"https://pith.science/api/pith-number/HR5W6PUJRZU6KPK7T5W2GOHUG5/graph.json","fetch_events":"https://pith.science/api/pith-number/HR5W6PUJRZU6KPK7T5W2GOHUG5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5/action/storage_attestation","attest_author":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5/action/author_attestation","sign_citation":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5/action/citation_signature","submit_replication":"https://pith.science/pith/HR5W6PUJRZU6KPK7T5W2GOHUG5/action/replication_record"}},"created_at":"2026-07-05T11:23:47.332936+00:00","updated_at":"2026-07-05T11:23:47.332936+00:00"}