{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:H6ZN4GPOK4AAV52YAPVIAEZZNH","short_pith_number":"pith:H6ZN4GPO","schema_version":"1.0","canonical_sha256":"3fb2de19ee57000af75803ea80133969d5c55c6e513e8718907ff207aa1f920e","source":{"kind":"arxiv","id":"2406.04770","version":2},"attestation_state":"computed","paper":{"title":"WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abhilasha Ravichander, Bill Yuchen Lin, Faeze Brahman, Khyathi Chandu, Nouha Dziri, Ronan Le Bras, Valentina Pyatkin, Yejin Choi, Yuntian Deng","submitted_at":"2024-06-07T09:15:44Z","abstract_excerpt":"We introduce WildBench, an automated evaluation framework designed to benchmark large language models (LLMs) using challenging, real-world user queries. WildBench consists of 1,024 tasks carefully selected from over one million human-chatbot conversation logs. For automated evaluation with WildBench, we have developed two metrics, WB-Reward and WB-Score, which are computable using advanced LLMs such as GPT-4-turbo. WildBench evaluation uses task-specific checklists to evaluate model outputs systematically and provides structured explanations that justify the scores and comparisons, resulting i"},"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":"2406.04770","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-07T09:15:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8b1fcbde56ae445804d87363a01e458cd3880bb6b1417693e7593ef2bc55695e","abstract_canon_sha256":"dbf5c8fdd875a506912cb1a7727e6ed4ab37f900644193aa1782a936230a65e2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:16:17.675844Z","signature_b64":"NHbtlwPsWXIV3FV1kU4loj0V6VQI89NvGyOz4QR1H9l4dzaG13Qy9mQV6/1noNMXETYEi05y1aIdF1Aa0gPSBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3fb2de19ee57000af75803ea80133969d5c55c6e513e8718907ff207aa1f920e","last_reissued_at":"2026-07-05T09:16:17.675253Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:16:17.675253Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abhilasha Ravichander, Bill Yuchen Lin, Faeze Brahman, Khyathi Chandu, Nouha Dziri, Ronan Le Bras, Valentina Pyatkin, Yejin Choi, Yuntian Deng","submitted_at":"2024-06-07T09:15:44Z","abstract_excerpt":"We introduce WildBench, an automated evaluation framework designed to benchmark large language models (LLMs) using challenging, real-world user queries. WildBench consists of 1,024 tasks carefully selected from over one million human-chatbot conversation logs. For automated evaluation with WildBench, we have developed two metrics, WB-Reward and WB-Score, which are computable using advanced LLMs such as GPT-4-turbo. WildBench evaluation uses task-specific checklists to evaluate model outputs systematically and provides structured explanations that justify the scores and comparisons, resulting i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.04770","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/2406.04770/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":"2406.04770","created_at":"2026-07-05T09:16:17.675340+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.04770v2","created_at":"2026-07-05T09:16:17.675340+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.04770","created_at":"2026-07-05T09:16:17.675340+00:00"},{"alias_kind":"pith_short_12","alias_value":"H6ZN4GPOK4AA","created_at":"2026-07-05T09:16:17.675340+00:00"},{"alias_kind":"pith_short_16","alias_value":"H6ZN4GPOK4AAV52Y","created_at":"2026-07-05T09:16:17.675340+00:00"},{"alias_kind":"pith_short_8","alias_value":"H6ZN4GPO","created_at":"2026-07-05T09:16:17.675340+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":20,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08758","citing_title":"Ideas Have Genomes: Benchmarking Scientific Lineage Reasoning and Lineage-Grounded Idea Generation","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2408.15549","citing_title":"WildFeedback: Aligning LLMs With In-situ User Interactions And Feedback","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18721","citing_title":"General Preference Reinforcement Learning","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18721","citing_title":"General Preference Reinforcement Learning","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20520","citing_title":"Open-World Evaluations for Measuring Frontier AI Capabilities","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21086","citing_title":"LoCar: Localization-Aware Evaluation of In-Vehicle Assistants through Fine-Grained Sociolinguistic Control","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18721","citing_title":"General Preference Reinforcement Learning","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2506.15787","citing_title":"SLR: Automated Synthesis for Scalable Logical Reasoning","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2509.11206","citing_title":"Evalet: Evaluating Large Language Models through Functional Fragmentation","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2603.06003","citing_title":"EvoESAP: Non-Uniform Expert Pruning for Sparse MoE","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2603.27977","citing_title":"SARL: Label-Free Reinforcement Learning by Rewarding Reasoning Topology","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2601.08584","citing_title":"Ministral 3","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09808","citing_title":"Quantifying the Utility of User Simulators for Building Collaborative LLM Assistants","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09228","citing_title":"ProactBench: Beyond What The User Asked For","ref_index":123,"is_internal_anchor":false},{"citing_arxiv_id":"2412.05579","citing_title":"LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods","ref_index":143,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07837","citing_title":"SPARD: Self-Paced Curriculum for RL Alignment via Integrating Reward Dynamics and Data Utility","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06996","citing_title":"Self-Preference Bias in Rubric-Based Evaluation of Large Language Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14585","citing_title":"Prompt Optimization Is a Coin Flip: Diagnosing When It Helps in Compound AI Systems","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04454","citing_title":"Deployment-Relevant Alignment Cannot Be Inferred from Model-Level Evaluation Alone","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02209","citing_title":"Submodular Benchmark Selection","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH","json":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH.json","graph_json":"https://pith.science/api/pith-number/H6ZN4GPOK4AAV52YAPVIAEZZNH/graph.json","events_json":"https://pith.science/api/pith-number/H6ZN4GPOK4AAV52YAPVIAEZZNH/events.json","paper":"https://pith.science/paper/H6ZN4GPO"},"agent_actions":{"view_html":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH","download_json":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH.json","view_paper":"https://pith.science/paper/H6ZN4GPO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.04770&json=true","fetch_graph":"https://pith.science/api/pith-number/H6ZN4GPOK4AAV52YAPVIAEZZNH/graph.json","fetch_events":"https://pith.science/api/pith-number/H6ZN4GPOK4AAV52YAPVIAEZZNH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH/action/storage_attestation","attest_author":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH/action/author_attestation","sign_citation":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH/action/citation_signature","submit_replication":"https://pith.science/pith/H6ZN4GPOK4AAV52YAPVIAEZZNH/action/replication_record"}},"created_at":"2026-07-05T09:16:17.675340+00:00","updated_at":"2026-07-05T09:16:17.675340+00:00"}