{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "zhang-et-al-infinitebench-2024", "title": "InfiniteBench: evaluation on contexts averaging beyond 100K tokens", "summary": "Combined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks.", "boundary": "The average exceeds 100K tokens; not every task or example does. Lengths and difficulty vary substantially, and input plus output lengths are separately reported. Capability is task- and metric-specific, not guaranteed for arbitrary 100K-token inputs.", "area": "Long-context evaluation", "topic": "Artificial intelligence", "year": 2024, "lane": "HUMAN", "result": "BENCHMARK", "contribution_kind": "EVALUATION_RESOURCE_DESIGN", "package_id": "zhang-et-al-infinitebench-2024", "target": { "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "content_hash": "sha256:bfe79c594714c1f64d028b09fcccc9337261bb7352c0f1546796a71d99a41bb8" }, "humans": [ "Xinrong Zhang", "Yingfa Chen", "Shengding Hu", "Zihang Xu", "Junhao Chen", "Moo Khai Hao", "Xu Han", "Zhen Leng Thai", "Shuo Wang", "Zhiyuan Liu", "Maosong Sun" ], "authors_complete": true, "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.", "agents": [], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "cur20261015-arxiv-2402-13718v1", "title": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } ], "points": "65000", "published_grade": { "id": "zhang-et-al-infinitebench-2024", "entry_id": "zhang-et-al-infinitebench-2024", "band": "B", "modifier": "+", "target": { "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "content_hash": "sha256:bfe79c594714c1f64d028b09fcccc9337261bb7352c0f1546796a71d99a41bb8" }, "what_changed": "Combined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks.", "scope": "Combined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks. The average exceeds 100K tokens; not every task or example does. Lengths and difficulty vary substantially, and input plus output lengths are separately reported. Capability is task- and metric-specific, not guaranteed for arbitrary 100K-token inputs.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "content_hash": "sha256:bfe79c594714c1f64d028b09fcccc9337261bb7352c0f1546796a71d99a41bb8" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-10-16T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [ { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Xinrong Zhang", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Yingfa Chen", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Shengding Hu", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zihang Xu", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Junhao Chen", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Moo Khai Hao", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Xu Han", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zhen Leng Thai", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Shuo Wang", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zhiyuan Liu", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Maosong Sun", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "The average exceeds 100K tokens; not every task or example does. Lengths and difficulty vary substantially, and input plus output lengths are separately reported. Capability is task- and metric-specific, not guaranteed for arbitrary 100K-token inputs.", "domain": "Long-context evaluation" }, "sources": [ { "citation": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1" } ], "statement": "# InfiniteBench: evaluation on contexts averaging beyond 100K tokens\n\nCombined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks.\n\n## Scientific scope\n\nThe average exceeds 100K tokens; not every task or example does. Lengths and difficulty vary substantially, and input plus output lengths are separately reported. Capability is task- and metric-specific, not guaranteed for arbitrary 100K-token inputs.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nXinrong Zhang; Yingfa Chen; Shengding Hu; Zihang Xu; Junhao Chen; Moo Khai Hao; Xu Han; Zhen Leng Thai; Shuo Wang; Zhiyuan Liu; Maosong Sun\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens](https://arxiv.org/abs/2402.13718v1)\n", "summary": "Combined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks.", "title": "InfiniteBench: evaluation on contexts averaging beyond 100K tokens" }, "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "source_ref": { "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "content_hash": "sha256:bfe79c594714c1f64d028b09fcccc9337261bb7352c0f1546796a71d99a41bb8" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:ec06df6198c51e2d0f471fdf3b0a997cbcaf1a4ef1a02240cca36cd4f1c0dc8a" }, "target": { "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "content_hash": "sha256:bfe79c594714c1f64d028b09fcccc9337261bb7352c0f1546796a71d99a41bb8" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "zhang-et-al-infinitebench-2024", "entry_id": "zhang-et-al-infinitebench-2024", "revision": 1, "title": "InfiniteBench: evaluation on contexts averaging beyond 100K tokens", "context": "Long-context evaluation", "track": "Evaluation resource design", "band": "B", "modifier": "+", "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-10-16", "evidence_cutoff": "2026-10-16", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "Combined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks.", "scope": "Combined synthetic and realistic English/Chinese tasks in retrieval, code, mathematics, novels and dialogue to test models beyond the usual short long-context benchmarks. The average exceeds 100K tokens; not every task or example does. Lengths and difficulty vary substantially, and input plus output lengths are separately reported. Capability is task- and metric-specific, not guaranteed for arbitrary 100K-token inputs.", "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. The average exceeds 100K tokens; not every task or example does. Lengths and difficulty vary substantially, and input plus output lengths are separately reported. Capability is task- and metric-specific, not guaranteed for arbitrary 100K-token inputs.", "references": [ { "title": "$\\infty$Bench: Extending Long Context Evaluation Beyond 100K Tokens", "url": "https://arxiv.org/abs/2402.13718v1", "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification." } ], "roles": [ { "who": "Source-listed authors and technical contributors", "credit": "SCIENCE", "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities." } ], "target": { "record_id": "urn:uuid:cb8f5af8-64ce-4611-a8a4-62f5cbfd53ce", "version": 1, "content_hash": "sha256:bfe79c594714c1f64d028b09fcccc9337261bb7352c0f1546796a71d99a41bb8" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "65000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": null, "reading_status": "SOURCE_LINKED_NOTE" }