{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XAIVPRJCP5WEPLFDWSCPFXBUT3","short_pith_number":"pith:XAIVPRJC","schema_version":"1.0","canonical_sha256":"b81157c5227f6c47aca3b484f2dc349ec3a0f5279aa27961c1a83395977477e8","source":{"kind":"arxiv","id":"2402.12170","version":3},"attestation_state":"computed","paper":{"title":"Where is the answer? Investigating Positional Bias in Language Model Knowledge Extraction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chen-Yu Lee, Kihyuk Sohn, Kuniaki Saito, Yoshitaka Ushiku","submitted_at":"2024-02-16T06:29:16Z","abstract_excerpt":"Large language models require updates to remain up-to-date or adapt to new domains by fine-tuning them with new documents. One key is memorizing the latest information in a way that the memorized information is extractable with a query prompt. However, LLMs suffer from a phenomenon called perplexity curse; despite minimizing document perplexity during fine-tuning, LLMs struggle to extract information through a prompt sentence. In this new knowledge acquisition and extraction, we find a very intriguing fact that LLMs can accurately answer questions about the first sentence, but they struggle to"},"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":"2402.12170","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-16T06:29:16Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"09d24eecede543f6b57e306caf996e135d0ff8696bca4f81f9ff3da94d5ac641","abstract_canon_sha256":"7b1a52f5e8cdf51d2f78e66792f66281ff17e70513b0b2bf49264a852c6d1a7e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:50:41.533823Z","signature_b64":"1sZXI8XHNVXAqbYXUqiAT4BoHHgx9DxBGO4ZNTCu4/JzC6vZLwRC3yoqWP9dCRzdItCYw/6R7aJ9yNI8rLr1DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b81157c5227f6c47aca3b484f2dc349ec3a0f5279aa27961c1a83395977477e8","last_reissued_at":"2026-07-05T10:50:41.533358Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:50:41.533358Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Where is the answer? Investigating Positional Bias in Language Model Knowledge Extraction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chen-Yu Lee, Kihyuk Sohn, Kuniaki Saito, Yoshitaka Ushiku","submitted_at":"2024-02-16T06:29:16Z","abstract_excerpt":"Large language models require updates to remain up-to-date or adapt to new domains by fine-tuning them with new documents. One key is memorizing the latest information in a way that the memorized information is extractable with a query prompt. However, LLMs suffer from a phenomenon called perplexity curse; despite minimizing document perplexity during fine-tuning, LLMs struggle to extract information through a prompt sentence. In this new knowledge acquisition and extraction, we find a very intriguing fact that LLMs can accurately answer questions about the first sentence, but they struggle to"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.12170","kind":"arxiv","version":3},"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/2402.12170/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":"2402.12170","created_at":"2026-07-05T10:50:41.533415+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.12170v3","created_at":"2026-07-05T10:50:41.533415+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.12170","created_at":"2026-07-05T10:50:41.533415+00:00"},{"alias_kind":"pith_short_12","alias_value":"XAIVPRJCP5WE","created_at":"2026-07-05T10:50:41.533415+00:00"},{"alias_kind":"pith_short_16","alias_value":"XAIVPRJCP5WEPLFD","created_at":"2026-07-05T10:50:41.533415+00:00"},{"alias_kind":"pith_short_8","alias_value":"XAIVPRJC","created_at":"2026-07-05T10:50:41.533415+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.12994","citing_title":"LogicEval: A Systematic Framework for Evaluating Automated Repair Techniques for Logical Vulnerabilities in Real-World Software","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3","json":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3.json","graph_json":"https://pith.science/api/pith-number/XAIVPRJCP5WEPLFDWSCPFXBUT3/graph.json","events_json":"https://pith.science/api/pith-number/XAIVPRJCP5WEPLFDWSCPFXBUT3/events.json","paper":"https://pith.science/paper/XAIVPRJC"},"agent_actions":{"view_html":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3","download_json":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3.json","view_paper":"https://pith.science/paper/XAIVPRJC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.12170&json=true","fetch_graph":"https://pith.science/api/pith-number/XAIVPRJCP5WEPLFDWSCPFXBUT3/graph.json","fetch_events":"https://pith.science/api/pith-number/XAIVPRJCP5WEPLFDWSCPFXBUT3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3/action/storage_attestation","attest_author":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3/action/author_attestation","sign_citation":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3/action/citation_signature","submit_replication":"https://pith.science/pith/XAIVPRJCP5WEPLFDWSCPFXBUT3/action/replication_record"}},"created_at":"2026-07-05T10:50:41.533415+00:00","updated_at":"2026-07-05T10:50:41.533415+00:00"}