{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:AGXSE5JWYX6K5OZRKBLN34PISS","short_pith_number":"pith:AGXSE5JW","schema_version":"1.0","canonical_sha256":"01af227536c5fcaebb315056ddf1e89493cb8619bfd522a35d52501c24df4372","source":{"kind":"arxiv","id":"2310.07641","version":2},"attestation_state":"computed","paper":{"title":"Evaluating Large Language Models at Evaluating Instruction Following","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Danqi Chen, Jiatong Yu, Tanya Goyal, Tianyu Gao, Yu Meng, Zhiyuan Zeng","submitted_at":"2023-10-11T16:38:11Z","abstract_excerpt":"As research in large language models (LLMs) continues to accelerate, LLM-based evaluation has emerged as a scalable and cost-effective alternative to human evaluations for comparing the ever increasing list of models. This paper investigates the efficacy of these ``LLM evaluators'', particularly in using them to assess instruction following, a metric that gauges how closely generated text adheres to the given instruction. We introduce a challenging meta-evaluation benchmark, LLMBar, designed to test the ability of an LLM evaluator in discerning instruction-following outputs. The authors manual"},"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":"2310.07641","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-11T16:38:11Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a2c12a534db8513c14260db501d325e3cfe97d9d4db39c4e64f337af874b7071","abstract_canon_sha256":"f9ffbe979d80f8ad85588003f7357a6278fbbf583726809f997c33d9f185df8b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:08:18.429248Z","signature_b64":"J2p0TqGzMOT+BewKpmiB1yJ6/Ss4vkUhBI/ooKeY1DmzYstJL1gk9NEeoj8Gsl6VmVRC8gt4kkQG7VZU+mSOCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01af227536c5fcaebb315056ddf1e89493cb8619bfd522a35d52501c24df4372","last_reissued_at":"2026-07-05T08:08:18.428785Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:08:18.428785Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluating Large Language Models at Evaluating Instruction Following","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Danqi Chen, Jiatong Yu, Tanya Goyal, Tianyu Gao, Yu Meng, Zhiyuan Zeng","submitted_at":"2023-10-11T16:38:11Z","abstract_excerpt":"As research in large language models (LLMs) continues to accelerate, LLM-based evaluation has emerged as a scalable and cost-effective alternative to human evaluations for comparing the ever increasing list of models. This paper investigates the efficacy of these ``LLM evaluators'', particularly in using them to assess instruction following, a metric that gauges how closely generated text adheres to the given instruction. We introduce a challenging meta-evaluation benchmark, LLMBar, designed to test the ability of an LLM evaluator in discerning instruction-following outputs. The authors manual"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.07641","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/2310.07641/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":"2310.07641","created_at":"2026-07-05T08:08:18.428844+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.07641v2","created_at":"2026-07-05T08:08:18.428844+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.07641","created_at":"2026-07-05T08:08:18.428844+00:00"},{"alias_kind":"pith_short_12","alias_value":"AGXSE5JWYX6K","created_at":"2026-07-05T08:08:18.428844+00:00"},{"alias_kind":"pith_short_16","alias_value":"AGXSE5JWYX6K5OZR","created_at":"2026-07-05T08:08:18.428844+00:00"},{"alias_kind":"pith_short_8","alias_value":"AGXSE5JW","created_at":"2026-07-05T08:08:18.428844+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19714","citing_title":"AURA: Adaptive Uncertainty-aware Refinement for LLM-as-a-Judge Auditing","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03102","citing_title":"Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time Scaling","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28365","citing_title":"CAMI: Cost-Aware Agent-Guided Multi-Indexing for Semantic Retrieval","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00093","citing_title":"Agreement Metrics for LLM-as-Judge Evaluation: What to Report and Why","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29430","citing_title":"Towards Human-Like Interactive Speech Recognition With Agentic Correction and Semantic Evaluation","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07006","citing_title":"RASFT: Rollout-Adaptive Supervised Fine-Tuning for Reasoning","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2410.04047","citing_title":"TS-Reasoner: Domain-Oriented Time Series Inference Agents for Reasoning and Automated Analysis","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2411.15594","citing_title":"A Survey on LLM-as-a-Judge","ref_index":202,"is_internal_anchor":false},{"citing_arxiv_id":"2404.13076","citing_title":"LLM Evaluators Recognize and Favor Their Own Generations","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19196","citing_title":"Time to REFLECT: Can We Trust LLM Judges for Evidence-based Research Agents?","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2408.07199","citing_title":"Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents","ref_index":172,"is_internal_anchor":false},{"citing_arxiv_id":"2507.06419","citing_title":"Teach a Reward Model to Correct Itself: Reward Guided Adversarial Failure Discovery for Robust Reward Modeling","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2410.18451","citing_title":"Skywork-Reward: Bag of Tricks for Reward Modeling in LLMs","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23178","citing_title":"Judging the Judges: A Systematic Evaluation of Bias Mitigation Strategies in LLM-as-a-Judge Pipelines","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS","json":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS.json","graph_json":"https://pith.science/api/pith-number/AGXSE5JWYX6K5OZRKBLN34PISS/graph.json","events_json":"https://pith.science/api/pith-number/AGXSE5JWYX6K5OZRKBLN34PISS/events.json","paper":"https://pith.science/paper/AGXSE5JW"},"agent_actions":{"view_html":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS","download_json":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS.json","view_paper":"https://pith.science/paper/AGXSE5JW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.07641&json=true","fetch_graph":"https://pith.science/api/pith-number/AGXSE5JWYX6K5OZRKBLN34PISS/graph.json","fetch_events":"https://pith.science/api/pith-number/AGXSE5JWYX6K5OZRKBLN34PISS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS/action/storage_attestation","attest_author":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS/action/author_attestation","sign_citation":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS/action/citation_signature","submit_replication":"https://pith.science/pith/AGXSE5JWYX6K5OZRKBLN34PISS/action/replication_record"}},"created_at":"2026-07-05T08:08:18.428844+00:00","updated_at":"2026-07-05T08:08:18.428844+00:00"}