{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:2RONHCQLISUZ6MQ6ZFASMWKRZQ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"0dab17f3bc82392eaaaef3fcf341d1b5e5128b248a791cf19655f7fd4814502c","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-20T03:26:47Z","title_canon_sha256":"5f1902dbfd7e2ea94af0f97fa55957f61ab2c8f47a8a7a59d3365df2a48bbb26"},"schema_version":"1.0","source":{"id":"2412.15524","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.15524","created_at":"2026-07-05T10:38:40Z"},{"alias_kind":"arxiv_version","alias_value":"2412.15524v2","created_at":"2026-07-05T10:38:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.15524","created_at":"2026-07-05T10:38:40Z"},{"alias_kind":"pith_short_12","alias_value":"2RONHCQLISUZ","created_at":"2026-07-05T10:38:40Z"},{"alias_kind":"pith_short_16","alias_value":"2RONHCQLISUZ6MQ6","created_at":"2026-07-05T10:38:40Z"},{"alias_kind":"pith_short_8","alias_value":"2RONHCQL","created_at":"2026-07-05T10:38:40Z"}],"graph_snapshots":[{"event_id":"sha256:6afb9f1f5e71ad2cfcbc695680c953122eae9608801188c00a2ce8b1432d6681","target":"graph","created_at":"2026-07-05T10:38:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2412.15524/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating the capability of Large Language Models (LLMs) in following instructions has heavily relied on a powerful LLM as the judge, introducing unresolved biases that deviate the judgments from human judges. In this work, we reevaluate various choices for automatic evaluation on a wide range of instruction-following tasks. We experiment with methods that leverage human-written responses and observe that they enhance the reliability of automatic evaluations across a wide range of tasks, resulting in up to a 3.2% improvement in agreement with human judges. We also discovered that human-writte","authors_text":"Hannaneh Hajishirzi, Pradeep Dasigi, Xinxi Lyu, Yizhong Wang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-20T03:26:47Z","title":"HREF: Human Response-Guided Evaluation of Instruction Following in Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.15524","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:cca357221cd5ac229bd84cf529e8ed4490e8a543599600504ec0979d87980e61","target":"record","created_at":"2026-07-05T10:38:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"0dab17f3bc82392eaaaef3fcf341d1b5e5128b248a791cf19655f7fd4814502c","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-20T03:26:47Z","title_canon_sha256":"5f1902dbfd7e2ea94af0f97fa55957f61ab2c8f47a8a7a59d3365df2a48bbb26"},"schema_version":"1.0","source":{"id":"2412.15524","kind":"arxiv","version":2}},"canonical_sha256":"d45cd38a0b44a99f321ec941265951cc35b1199539cd8e57118d82c48315f8de","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d45cd38a0b44a99f321ec941265951cc35b1199539cd8e57118d82c48315f8de","first_computed_at":"2026-07-05T10:38:40.532587Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:38:40.532587Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cwmjVD7jPCbqk73blqb947PUJd70Ca4C3HHyx+OPAMz+Y+M+f8F1JuAar91r7zeA71asjHrXE6PIwHAH83/RDg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:38:40.533083Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.15524","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:cca357221cd5ac229bd84cf529e8ed4490e8a543599600504ec0979d87980e61","sha256:6afb9f1f5e71ad2cfcbc695680c953122eae9608801188c00a2ce8b1432d6681"],"state_sha256":"cb43b67482be20b439463b7aed4ada145a35bb8ec79894fbc568b145ef444ab0"}