{"record": {"aliases": ["Measuring in-context algorithmic reasoning in language models against an exact Bayes-optimal standard"], "artifacts": [{"id": "arxiv:2608.01575", "kind": "paper", "url": "https://arxiv.org/abs/2608.01575"}], "authors": ["Luan Ozelim", "Hector Zenil"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.01575", "source_url": "https://arxiv.org/abs/2608.01575"}, "categories": ["General AI", "Language & Knowledge", "Reasoning"], "description": {"en": "F-ICL evaluates in-context algorithmic inference against an exactly computed Bayes-optimal posterior under a bounded universal prior."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.01575", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.01575", "title": "F-ICL"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["King's College London", "King's College School"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2608.01575", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2608.01575", "source_url": "https://openalex.org/W7172437980"}, "key": "claire-radar:2608.01575", "modality": null, "name": "F-ICL", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_measuring-in-context-algorithmic-reasoning_e1942bef", "origin_source": "arxiv", "recognition_confidence": "0.6", "record_sha256": "f45a2ddca2f7fcb9a365b35b132e09a2810c9e40651de20a7c5212b2a407e719", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T19:30:29.387990Z", "source_url": "https://arxiv.org/abs/2608.01575"}, "publisher": null, "released": "2026-08-03", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.01575", "source_url": "https://arxiv.org/abs/2608.01575"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-01575", "source": "claire_radar", "source_benchmark_id": "2608.01575", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Measuring in-context algorithmic reasoning in language models against an exact Bayes-optimal standard"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The exact Bayes-optimal reference and broad evaluation across open and frontier models may attract research attention.", "score": 7}, "capabilities": ["Reasoning"], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T19:30:29.387990Z", "inputHash": "78c3d8551f83d39d7aff2d6e54449524528e47e95f29f4df33dbeeaac3c2409c", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "We introduce F-ICL, an in-context-learning benchmark", "canonicalNameSource": "abstract", "decisionReason": "The source formally names the benchmark, states a repeatable task and comparable scoring rule, and explicitly says the benchmark and toolkit are openly released.", "evaluationEvidence": {"reuse": "F-ICL is released as an open, reproducible benchmark and toolkit.", "scoring": "models are scored by how closely their served distributions approach it at matched evidence", "task": "an in-context-learning benchmark that supplies one exactly"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T19:30:29.387990Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "F-ICL evaluates in-context algorithmic inference against an exactly computed Bayes-optimal posterior under a bounded universal prior.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce F-ICL, an in-context-learning benchmark that supplies one exactly."}, "familyId": "bmf_98eb8416ff54", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_measuring-in-context-algorithmic-reasoning_e1942bef", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2608.01575", "pdf": "https://arxiv.org/pdf/2608.01575", "project": null, "report": "https://arxiv.org/abs/2608.01575"}, "metrics": [], "motivation": "Whether large language models perform genuine algorithmic reasoning or mere pattern completion is hard to test, because most benchmarks lack a ground truth for correct inductive inference.", "name": "F-ICL", "oneLine": "F-ICL evaluates in-context algorithmic inference against an exactly computed Bayes-optimal posterior under a bounded universal prior.", "paperTitle": "Measuring in-context algorithmic reasoning in language models against an exact Bayes-optimal standard", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.01575", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Paper only", "recognitionConfidence": 0.6, "relation": "introduces", "releasedAt": "2026-08-03", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Hector Zenil and Luan Ozelim"], "categories": ["cs.LG", "cs.AI"], "id": "2608.01575", "title": "Measuring in-context algorithmic reasoning in language models against an exact Bayes-optimal standard", "type": "arxiv", "url": "https://arxiv.org/abs/2608.01575"}, "sourceUpdatedAt": "2026-08-03T01:21:44Z", "topics": ["Reasoning"], "venueAttempts": [], "whyItMatters": "It provides an objective reference for separating algorithmic inference from simpler pattern-based behavior."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
