{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6ZEZN7YMAYB46GMILRMNL64C3K","short_pith_number":"pith:6ZEZN7YM","schema_version":"1.0","canonical_sha256":"f64996ff0c0603cf19885c58d5fb82da99cb0d67a7bc5fff45701ae52a94cd5e","source":{"kind":"arxiv","id":"2404.09971","version":2},"attestation_state":"computed","paper":{"title":"Constructing Benchmarks and Interventions for Combating Hallucinations in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adi Simhi, Idan Szpektor, Jonathan Herzig, Yonatan Belinkov","submitted_at":"2024-04-15T17:48:46Z","abstract_excerpt":"Large language models (LLMs) are prone to hallucinations, which sparked a widespread effort to detect and prevent them. Recent work attempts to mitigate hallucinations by intervening in the model's generation, typically computing representative vectors of hallucinations vs. grounded generations, for steering the model's hidden states away from a hallucinatory state. However, common studies employ different setups and do not properly separate different possible causes of hallucinations, making interventions misguided. In this work, we introduce a method for categorizing examples based on the mo"},"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":"2404.09971","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-04-15T17:48:46Z","cross_cats_sorted":[],"title_canon_sha256":"94a34ee2d482deeb946c90081f62408ccc8ea3c4413ad46e6c15835227c94477","abstract_canon_sha256":"363f0bc3a37f85788ae06badf0e1b3e74659e978d3231d478f02194c77a34f17"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:42:35.168189Z","signature_b64":"URZHu4Q1h7wwhYY7CT6H85sIBKrbv/fd3hcSHWv0EPpNXilwWYyxMMgBFZrWKBwrwjN64EudZvfz2DrK43xzDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f64996ff0c0603cf19885c58d5fb82da99cb0d67a7bc5fff45701ae52a94cd5e","last_reissued_at":"2026-07-05T08:42:35.167700Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:42:35.167700Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Constructing Benchmarks and Interventions for Combating Hallucinations in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adi Simhi, Idan Szpektor, Jonathan Herzig, Yonatan Belinkov","submitted_at":"2024-04-15T17:48:46Z","abstract_excerpt":"Large language models (LLMs) are prone to hallucinations, which sparked a widespread effort to detect and prevent them. Recent work attempts to mitigate hallucinations by intervening in the model's generation, typically computing representative vectors of hallucinations vs. grounded generations, for steering the model's hidden states away from a hallucinatory state. However, common studies employ different setups and do not properly separate different possible causes of hallucinations, making interventions misguided. In this work, we introduce a method for categorizing examples based on the mo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.09971","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/2404.09971/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":"2404.09971","created_at":"2026-07-05T08:42:35.167760+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.09971v2","created_at":"2026-07-05T08:42:35.167760+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.09971","created_at":"2026-07-05T08:42:35.167760+00:00"},{"alias_kind":"pith_short_12","alias_value":"6ZEZN7YMAYB4","created_at":"2026-07-05T08:42:35.167760+00:00"},{"alias_kind":"pith_short_16","alias_value":"6ZEZN7YMAYB46GMI","created_at":"2026-07-05T08:42:35.167760+00:00"},{"alias_kind":"pith_short_8","alias_value":"6ZEZN7YM","created_at":"2026-07-05T08:42:35.167760+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09255","citing_title":"RPO-PDT: Demonstrating Role-Play-Based Knowledge Adaptation for Student Support Dialogue (Demonstration System)","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05054","citing_title":"Boosting Self-Consistency with Ranking","ref_index":245,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24957","citing_title":"Mitigating Object Hallucinations in Vision-Language Models through Region-Aware Attention Recalibration","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05244","citing_title":"Towards Dependable Retrieval-Augmented Generation Using Factual Confidence Prediction","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K","json":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K.json","graph_json":"https://pith.science/api/pith-number/6ZEZN7YMAYB46GMILRMNL64C3K/graph.json","events_json":"https://pith.science/api/pith-number/6ZEZN7YMAYB46GMILRMNL64C3K/events.json","paper":"https://pith.science/paper/6ZEZN7YM"},"agent_actions":{"view_html":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K","download_json":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K.json","view_paper":"https://pith.science/paper/6ZEZN7YM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.09971&json=true","fetch_graph":"https://pith.science/api/pith-number/6ZEZN7YMAYB46GMILRMNL64C3K/graph.json","fetch_events":"https://pith.science/api/pith-number/6ZEZN7YMAYB46GMILRMNL64C3K/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K/action/storage_attestation","attest_author":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K/action/author_attestation","sign_citation":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K/action/citation_signature","submit_replication":"https://pith.science/pith/6ZEZN7YMAYB46GMILRMNL64C3K/action/replication_record"}},"created_at":"2026-07-05T08:42:35.167760+00:00","updated_at":"2026-07-05T08:42:35.167760+00:00"}