{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:OVRPLXFFPAMB3YJMZVR7RN7HJT","short_pith_number":"pith:OVRPLXFF","schema_version":"1.0","canonical_sha256":"7562f5dca578181de12ccd63f8b7e74cce35c88a0a6e426f2bb8d5305dc1788d","source":{"kind":"arxiv","id":"2402.10646","version":2},"attestation_state":"computed","paper":{"title":"AbsInstruct: Eliciting Abstraction Ability from LLMs through Explanation Tuning with Plausibility Estimation","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ginny Y. Wong, Hongming Zhang, Qing Zong, Sehyun Choi, Simon See, Tianqing Fang, Wei Fan, Xin Liu, Yangqiu Song, Zhaowei Wang","submitted_at":"2024-02-16T12:47:11Z","abstract_excerpt":"Abstraction ability is crucial in human intelligence, which can also benefit various tasks in NLP study. Existing work shows that LLMs are deficient in abstract ability, and how to improve it remains unexplored. In this work, we design the framework AbsInstruct to enhance LLMs' abstraction ability through instruction tuning. The framework builds instructions with in-depth explanations to assist LLMs in capturing the underlying rationale of abstraction. Meanwhile, we introduce a plausibility estimator to select instructions that are more consistent with the abstraction knowledge of LLMs to be a"},"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":"2402.10646","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-16T12:47:11Z","cross_cats_sorted":[],"title_canon_sha256":"f52c2d555d0691c6d5aefcb339a207a7d74371bd3225cc321e9ff335fd059f19","abstract_canon_sha256":"6e2272c8e40876020d789bae0f8e1b947fce1cc8c334ef4d0e90620de98fa866"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:32:36.781865Z","signature_b64":"s+9nvi/8x/UhuYAKPZ3TCcCxhsaREFFfCQXzhqy/0xUL/X/shpXDHJFNdLXE+Zp0NMt+8eSisDTaXIMDvBHpAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7562f5dca578181de12ccd63f8b7e74cce35c88a0a6e426f2bb8d5305dc1788d","last_reissued_at":"2026-07-05T08:32:36.781321Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:32:36.781321Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AbsInstruct: Eliciting Abstraction Ability from LLMs through Explanation Tuning with Plausibility Estimation","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ginny Y. Wong, Hongming Zhang, Qing Zong, Sehyun Choi, Simon See, Tianqing Fang, Wei Fan, Xin Liu, Yangqiu Song, Zhaowei Wang","submitted_at":"2024-02-16T12:47:11Z","abstract_excerpt":"Abstraction ability is crucial in human intelligence, which can also benefit various tasks in NLP study. Existing work shows that LLMs are deficient in abstract ability, and how to improve it remains unexplored. In this work, we design the framework AbsInstruct to enhance LLMs' abstraction ability through instruction tuning. The framework builds instructions with in-depth explanations to assist LLMs in capturing the underlying rationale of abstraction. Meanwhile, we introduce a plausibility estimator to select instructions that are more consistent with the abstraction knowledge of LLMs to be a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.10646","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/2402.10646/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":"2402.10646","created_at":"2026-07-05T08:32:36.781381+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.10646v2","created_at":"2026-07-05T08:32:36.781381+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.10646","created_at":"2026-07-05T08:32:36.781381+00:00"},{"alias_kind":"pith_short_12","alias_value":"OVRPLXFFPAMB","created_at":"2026-07-05T08:32:36.781381+00:00"},{"alias_kind":"pith_short_16","alias_value":"OVRPLXFFPAMB3YJM","created_at":"2026-07-05T08:32:36.781381+00:00"},{"alias_kind":"pith_short_8","alias_value":"OVRPLXFF","created_at":"2026-07-05T08:32:36.781381+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.19996","citing_title":"ReSURE: Regularizing Supervision Unreliability for Multi-turn Dialogue Fine-tuning","ref_index":57,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT","json":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT.json","graph_json":"https://pith.science/api/pith-number/OVRPLXFFPAMB3YJMZVR7RN7HJT/graph.json","events_json":"https://pith.science/api/pith-number/OVRPLXFFPAMB3YJMZVR7RN7HJT/events.json","paper":"https://pith.science/paper/OVRPLXFF"},"agent_actions":{"view_html":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT","download_json":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT.json","view_paper":"https://pith.science/paper/OVRPLXFF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.10646&json=true","fetch_graph":"https://pith.science/api/pith-number/OVRPLXFFPAMB3YJMZVR7RN7HJT/graph.json","fetch_events":"https://pith.science/api/pith-number/OVRPLXFFPAMB3YJMZVR7RN7HJT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT/action/storage_attestation","attest_author":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT/action/author_attestation","sign_citation":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT/action/citation_signature","submit_replication":"https://pith.science/pith/OVRPLXFFPAMB3YJMZVR7RN7HJT/action/replication_record"}},"created_at":"2026-07-05T08:32:36.781381+00:00","updated_at":"2026-07-05T08:32:36.781381+00:00"}