{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.12404", "kind": "paper", "url": "https://arxiv.org/abs/2609.12404"}], "authors": ["Yu Bai", "Yukai Miao", "Dawei Wang", "Li Chen", "Yanyu Ren", "Yuqian Shi", "Dan Li", "Ying Xiong", "Chengqiu Tan", "Run Zhou", "Li Li"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.12404", "source_url": "https://arxiv.org/abs/2609.12404"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "Evaluates agents on computer control tasks under finite trial budgets."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.12404", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.12404", "title": "VRL-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Tsinghua University", "China Mobile (China)"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2609.12404", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.12404", "source_url": "https://openalex.org/W7212923777"}, "key": "claire-radar:2609.12404", "modality": null, "name": "VRL-Bench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "false", "origin_record_id": "bm_vrl-bench_58e2cbeb", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "2ae976d5c542e93c5e7302ef8d88e0a4cbb36ed60d7490809367d220c24eb398", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-19T08:55:40.500135Z", "source_url": "https://arxiv.org/abs/2609.12404"}, "publisher": null, "released": "2026-09-11", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.12404", "source_url": "https://arxiv.org/abs/2609.12404"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-12404", "source": "claire_radar", "source_benchmark_id": "2609.12404", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-09-14", "attemptedAt": "2026-09-14T01:54:58.773663Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.12404", "observedAt": "2026-09-13T12:53:02.693952Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-14T01:54:58.773663Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The title and abstract describe a niche fair-evaluation harness for agent trial budgets without supplied artifacts or community signals.", "score": 10}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-19T08:55:40.500135Z", "inputHash": "1c4d711370172d3b99bb3ede457d545845681b4a9c550077c42cd9830e0cf4c3", "model": "deepseek-v4-pro"}, "curation": {"decisionReason": "No evidence is supplied for task instances, score definition, or public release path beyond the paper text.", "model": "deepseek-v4-pro", "policy": "source-evidence-v2", "reviewedAt": "2026-09-19T08:55:40.500135Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates agents on computer control tasks under finite trial budgets.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce VRL-Bench, a harness for fair evaluation of trial-and-error learning under finite trial budgets."}, "familyId": "bmf_7df8fc3cbc93", "firstSeenAt": "2026-09-14", "heat": null, "id": "bm_vrl-bench_58e2cbeb", "indexedAt": "2026-09-14T06:39:43.852630Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.12404", "pdf": "https://arxiv.org/pdf/2609.12404", "project": null, "report": "https://arxiv.org/abs/2609.12404"}, "metrics": [], "motivation": "Learning from trial and error is a promising way to improve language agents on complex tasks such as computer control.", "name": "VRL-Bench", "oneLine": "Evaluates agents on computer control tasks under finite trial budgets.", "paperTitle": "VRL-Bench: Benchmarking agents on computer control tasks under finite trial budgets", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.12404", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-14T06:39:43.852630Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}, "90d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-11", "source": {"authors": ["Yu Bai", "Yukai Miao", "Dawei Wang", "Li Chen", "Yanyu Ren", "Yuqian Shi", "Dan Li", "Ying Xiong", "Chengqiu Tan", "Run Zhou", "Li Li"], "categories": ["cs.AI"], "id": "2609.12404", "title": "VRL-Bench: Benchmarking agents on computer control tasks under finite trial budgets", "type": "arxiv", "url": "https://arxiv.org/abs/2609.12404"}, "sourceUpdatedAt": "2026-09-11T03:54:02Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Assesses trial-and-error learning under limited interaction budgets."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
