{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4TKVYUQ35CIMP6EFZW2EEDAWDD","short_pith_number":"pith:4TKVYUQ3","schema_version":"1.0","canonical_sha256":"e4d55c521be890c7f885cdb4420c1618eac1c35bef825b35821ce0dc4a4c10b3","source":{"kind":"arxiv","id":"2604.17237","version":2},"attestation_state":"computed","paper":{"title":"HeadRank: Decoding-Free Passage Reranking via Preference-Aligned Attention Heads","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"HeadRank reranks passages by aligning LLM attention heads to preferences in continuous space without decoding.","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Aolin Li, Chenxing Wang, Dongliang Liao, Haijun Wu, Huiyun Hu, Jin Xu, Junwu Du, Juyuan Wang, Ligang Liu, Shunlin Rong, Yuchen Fang","submitted_at":"2026-04-19T03:43:42Z","abstract_excerpt":"Decoding-free reranking methods that read relevance signals directly from LLM attention weights offer significant latency advantages over autoregressive approaches, yet suffer from attention score homogenization: middle-context documents receive near-identical scores, destroying the fine-grained distinctions required for ranking. We propose HeadRank, a framework that lifts preference optimization from discrete token space into the continuous attention domain through entropy-regularized head selection, hard adjacent-level preference pairs, and a distribution regularizer that jointly sharpen dis"},"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":"2604.17237","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2026-04-19T03:43:42Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"460dfc3d414688b56310e270bb82f31d4742ed23010bf53b87b7a06001d8e8ef","abstract_canon_sha256":"44ad9b279e33eb1e6c26b522cd6cda643171e9178f61986c6cf3972ea785d4e9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:13.894354Z","signature_b64":"q3WmNBLVeCyc89lUY0OAweY6bYWn8UqHWSkgbwrOC8N8AeZJdCkxkG57VbYj9L6YjwQKQQlIfpYfVyKbOhipDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4d55c521be890c7f885cdb4420c1618eac1c35bef825b35821ce0dc4a4c10b3","last_reissued_at":"2026-05-20T01:05:13.893425Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:13.893425Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HeadRank: Decoding-Free Passage Reranking via Preference-Aligned Attention Heads","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"HeadRank reranks passages by aligning LLM attention heads to preferences in continuous space without decoding.","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Aolin Li, Chenxing Wang, Dongliang Liao, Haijun Wu, Huiyun Hu, Jin Xu, Junwu Du, Juyuan Wang, Ligang Liu, Shunlin Rong, Yuchen Fang","submitted_at":"2026-04-19T03:43:42Z","abstract_excerpt":"Decoding-free reranking methods that read relevance signals directly from LLM attention weights offer significant latency advantages over autoregressive approaches, yet suffer from attention score homogenization: middle-context documents receive near-identical scores, destroying the fine-grained distinctions required for ranking. We propose HeadRank, a framework that lifts preference optimization from discrete token space into the continuous attention domain through entropy-regularized head selection, hard adjacent-level preference pairs, and a distribution regularizer that jointly sharpen dis"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"HeadRank consistently outperforms generative and decoding-free baselines with 100% formatting success. At 4B, 57.4% of relevant middle-zone documents reach the top quartile versus 14.2% for irrelevant ones -- a 43-percentage-point selectivity gap that demonstrates the effectiveness of attention-space preference alignment for listwise reranking.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That entropy-regularized head selection combined with hard adjacent-level preference pairs and a distribution regularizer can reliably overcome attention homogenization in middle context using only 211 training queries without introducing new biases or overfitting to the small training set.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"HeadRank improves decoding-free passage reranking by preference-aligning attention heads to increase discriminability in middle-context documents, outperforming baselines on 14 benchmarks with only 211 training queries.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"HeadRank reranks passages by aligning LLM attention heads to preferences in continuous space without decoding.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"6ba0c05f1268c983e42c044331acbab37d2e608a6661f2f2f869a04e2edd3fb2"},"source":{"id":"2604.17237","kind":"arxiv","version":2},"verdict":{"id":"873dd888-bab8-4e3b-8f88-e5afebace4d0","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T06:31:49.425911Z","strongest_claim":"HeadRank consistently outperforms generative and decoding-free baselines with 100% formatting success. At 4B, 57.4% of relevant middle-zone documents reach the top quartile versus 14.2% for irrelevant ones -- a 43-percentage-point selectivity gap that demonstrates the effectiveness of attention-space preference alignment for listwise reranking.","one_line_summary":"HeadRank improves decoding-free passage reranking by preference-aligning attention heads to increase discriminability in middle-context documents, outperforming baselines on 14 benchmarks with only 211 training queries.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That entropy-regularized head selection combined with hard adjacent-level preference pairs and a distribution regularizer can reliably overcome attention homogenization in middle context using only 211 training queries without introducing new biases or overfitting to the small training set.","pith_extraction_headline":"HeadRank reranks passages by aligning LLM attention heads to preferences in continuous space without decoding."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.17237/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":"2604.17237","created_at":"2026-05-20T01:05:13.893575+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.17237v2","created_at":"2026-05-20T01:05:13.893575+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.17237","created_at":"2026-05-20T01:05:13.893575+00:00"},{"alias_kind":"pith_short_12","alias_value":"4TKVYUQ35CIM","created_at":"2026-05-20T01:05:13.893575+00:00"},{"alias_kind":"pith_short_16","alias_value":"4TKVYUQ35CIMP6EF","created_at":"2026-05-20T01:05:13.893575+00:00"},{"alias_kind":"pith_short_8","alias_value":"4TKVYUQ3","created_at":"2026-05-20T01:05:13.893575+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD","json":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD.json","graph_json":"https://pith.science/api/pith-number/4TKVYUQ35CIMP6EFZW2EEDAWDD/graph.json","events_json":"https://pith.science/api/pith-number/4TKVYUQ35CIMP6EFZW2EEDAWDD/events.json","paper":"https://pith.science/paper/4TKVYUQ3"},"agent_actions":{"view_html":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD","download_json":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD.json","view_paper":"https://pith.science/paper/4TKVYUQ3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.17237&json=true","fetch_graph":"https://pith.science/api/pith-number/4TKVYUQ35CIMP6EFZW2EEDAWDD/graph.json","fetch_events":"https://pith.science/api/pith-number/4TKVYUQ35CIMP6EFZW2EEDAWDD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD/action/storage_attestation","attest_author":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD/action/author_attestation","sign_citation":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD/action/citation_signature","submit_replication":"https://pith.science/pith/4TKVYUQ35CIMP6EFZW2EEDAWDD/action/replication_record"}},"created_at":"2026-05-20T01:05:13.893575+00:00","updated_at":"2026-05-20T01:05:13.893575+00:00"}