{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.24882", "kind": "paper", "url": "https://arxiv.org/abs/2607.24882"}, {"id": "gh:eyuansu62/agent-retrieval-bench", "kind": "repo", "url": "https://github.com/eyuansu62/agent-retrieval-bench"}], "authors": ["Bowen Qin", "Yi Xie"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.24882", "source_url": "https://arxiv.org/abs/2607.24882"}, "categories": ["Agents", "Information retrieval", "Language & Knowledge", "Software & AI Compute"], "description": {"en": "File-level retrieval benchmark for coding agents, covering four positive tasks (code2test, comment2context, trace2code, edit2ripple) and a selective-retrieval subset with natural no-gold and counterfactual controls across 25 repositories, 427 samples, with frozen base-commit corpora."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.24882", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.24882", "title": "Agent Retrieval Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["National University of Singapore", "Peking University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.24882", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.24882", "source_url": "https://openalex.org/W7171670812"}, "key": "claire-radar:2607.24882", "modality": null, "name": "Agent Retrieval 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": "unknown", "origin_record_id": "bm_agent-retrieval-bench_1105bc27", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "affd8e04124e805d6136d14d929cab1e2d99d5923f16c148a43db05102468335", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.24882"}, "publisher": null, "released": "2026-07-27", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.24882", "source_url": "https://arxiv.org/abs/2607.24882"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-24882", "source": "claire_radar", "source_benchmark_id": "2607.24882", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Software & AI Compute"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/eyuansu62/agent-retrieval-bench", "githubScope": "benchmark_repo", "githubStars": 9, "hfDailySubmittedAt": "2026-07-29T00:00:00.000Z", "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 4, "hfPaperUrl": "https://huggingface.co/papers/2607.24882", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "capabilities": ["Information retrieval"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "4729577774258c87bd81952797c80bdace3b32993b4c2e0c7e1075a47f3acc68", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "File-level retrieval benchmark for coding agents, covering four positive tasks (code2test, comment2context, trace2code, edit2ripple) and a selective-retrieval subset with natural no-gold and counterfactual controls across 25 repositories, 427 samples, with frozen base-commit corpora.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce Agent Retrieval Bench, a file-level benchmark for this upstream retrieval problem."}, "familyId": "bmf_0f8de99c2972", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/eyuansu62/agent-retrieval-bench", "url": "https://github.com/eyuansu62/agent-retrieval-bench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/eyuansu62/agent-retrieval-bench", "url": "https://huggingface.co/papers/2607.24882"}, {"basis": "huggingface-paper-association", "linkedUrl": "https://github.com/eyuansu62/agent-retrieval-bench", "url": "https://huggingface.co/api/papers/2607.24882"}], "scope": "benchmark_repo", "url": "https://github.com/eyuansu62/agent-retrieval-bench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_agent-retrieval-bench_1105bc27", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Software & Cloud"], "links": {"code": "https://github.com/eyuansu62/agent-retrieval-bench", "data": null, "hfPaper": "https://huggingface.co/papers/2607.24882", "paper": "https://arxiv.org/abs/2607.24882", "pdf": "https://arxiv.org/pdf/2607.24882", "project": null, "report": "https://arxiv.org/abs/2607.24882"}, "metrics": [], "motivation": "Modern coding agents are usually evaluated by whether they eventually produce a correct patch, but patch generation depends on an earlier context-acquisition stage: finding the repository files needed for the task.", "name": "Agent Retrieval Bench", "oneLine": "File-level retrieval benchmark for coding agents, covering four positive tasks (code2test, comment2context, trace2code, edit2ripple) and a selective-retrieval subset with natural no-gold and counterfactual controls across 25 repositories, 427 samples, with frozen base-commit corpora.", "paperTitle": "Agent Retrieval Bench: Evaluating Repository Context Retrieval for Coding Agents", "primaryDomain": "Software & AI Compute", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.24882", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "Agent Retrieval Bench team", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/eyuansu62/agent-retrieval-bench"}], "ranking": {"90d": {"confidence": "Medium", "coverage": 0.7, "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": 0.30735371836622627, "value": 9}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": 0.2843790034276272, "value": 4}}, "confidence": "Medium", "coverage": 0.7, "rank": 170, "score": 36}, "method": "attention-ranking-v13", "rank": 170, "score": 36}}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-27", "source": {"authors": ["Bowen Qin", "Yi Xie"], "categories": ["cs.IR", "cs.AI", "cs.CL"], "id": "2607.24882", "title": "Agent Retrieval Bench: Evaluating Repository Context Retrieval for Coding Agents", "type": "arxiv", "url": "https://arxiv.org/abs/2607.24882"}, "sourceUpdatedAt": "2026-07-27T09:39:09Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "Provides a dedicated evaluation for the context-acquisition stage of coding agents, distinguishing retrieval quality from patch generation and offering a reusable protocol for comparing retrieval and selective abstention methods."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
