{"record": {"aliases": ["ComplexConstraints and Beyond: Expert Rubrics for RLVR"], "artifacts": [{"id": "arxiv:2606.09118", "kind": "paper", "url": "https://arxiv.org/abs/2606.09118"}, {"kind": "website", "url": "https://gem-workshop.com/"}], "authors": ["Sushant Mehta", "Liudas Panavas", "Suhaas Garre", "Edwin Chen"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.09118", "source_url": "https://arxiv.org/abs/2606.09118"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "An expert-curated instruction-following benchmark containing 75 public prompts and 1,559 atomic rubric criteria."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.09118", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.09118", "title": "ComplexConstraints"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.09118", "modality": null, "name": "ComplexConstraints", "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_complexconstraints-and-beyond-expert-rubri_f5e8b98d", "origin_source": "arxiv", "recognition_confidence": "0.7", "record_sha256": "4ece201220c0a4d3d6ef46fd2bc5fc6a57a70cb132e987066aff29bd4f5c97aa", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T17:05:39.521708Z", "source_url": "https://arxiv.org/abs/2606.09118"}, "publisher": null, "released": "2026-06-08", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.09118", "source_url": "https://arxiv.org/abs/2606.09118"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-09118", "source": "claire_radar", "source_benchmark_id": "2606.09118", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["ComplexConstraints and Beyond: Expert Rubrics for RLVR"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The public rubric-based evaluation suite may attract near-term interest from instruction-following and RLVR researchers.", "score": 3}, "capabilities": [], "confidence": "Medium", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T17:05:39.521708Z", "inputHash": "f46b72fbc77bf295c9ef526423846805cf9fe356001a0c3e92a59e9abcb48fef", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "We introduce ComplexConstraints", "canonicalNameSource": "abstract", "decisionReason": "The abstract formally names the benchmark, defines its evaluation task and criterion pass-rate scoring, and identifies a public benchmark release.", "evaluationEvidence": {"reuse": "a public 75-prompt benchmark with 1,559 rubric criteria", "scoring": "mean criterion pass rate", "task": "complex instruction following"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T17:05:39.521708Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "An expert-curated instruction-following benchmark containing 75 public prompts and 1,559 atomic rubric criteria.", "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", "public artifact URL"], "snippet": "We introduce ComplexConstraints, an expert-curated instruction-following suite comprising a public 75-prompt benchmark with 1,559 rubric criteria and a disjoint 1,000-prompt training set, with 10-40 atomic criteria per prompt."}, "familyId": "bmf_5b62063670bd", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_complexconstraints-and-beyond-expert-rubri_f5e8b98d", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2606.09118", "pdf": "https://arxiv.org/pdf/2606.09118", "project": "https://gem-workshop.com/", "report": "https://arxiv.org/abs/2606.09118"}, "metrics": [], "motivation": "Evaluation protocols can lag behind LLM capabilities.", "name": "ComplexConstraints", "oneLine": "An expert-curated instruction-following benchmark containing 75 public prompts and 1,559 atomic rubric criteria.", "paperTitle": "ComplexConstraints and Beyond: Expert Rubrics for RLVR", "primaryDomain": "General AI", "publication": {"evidence": "Accepted to the GEM workshop at ACL 2026: https://gem-workshop.com/", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2606.09118", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "GEM workshop at ACL 2026: https://gem-workshop", "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Paper only", "recognitionConfidence": 0.7, "relation": "introduces", "releasedAt": "2026-06-08", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Sushant Mehta", "Liudas Panavas", "Suhaas Garre", "Edwin Chen"], "categories": ["cs.AI"], "id": "2606.09118", "title": "ComplexConstraints and Beyond: Expert Rubrics for RLVR", "type": "arxiv", "url": "https://arxiv.org/abs/2606.09118"}, "sourceUpdatedAt": "2026-07-04T02:42:19Z", "topics": ["cs.AI"], "venueAttempts": [{"decisionRaw": "Accepted to the GEM workshop at ACL 2026: https://gem-workshop.com/", "evidence": [{"level": "author-claim", "observedAt": "2026-08-19T11:05:13.395391Z", "rawValue": "Accepted to the GEM workshop at ACL 2026: https://gem-workshop.com/", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2606.09118"}], "reviewStatus": "accepted", "venueName": "GEM workshop at ACL 2026: https://gem-workshop"}], "whyItMatters": "It supports criterion-level evaluation of semantic, contextual, and policy-dependent instruction following beyond narrow programmatic checks."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
