{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.21217", "kind": "paper", "url": "https://arxiv.org/abs/2607.21217"}, {"id": "gh:alex-nlp/icae-eval", "kind": "repo", "url": "https://github.com/ALEX-nlp/ICAE-EVAL"}], "authors": ["Zhongyuan Peng", "Dan Huang", "Chuyu Zhang", "Caijun Xu", "Changyi Xiao", "Shibo Hong", "David Lo", "Lin Qiu", "Xuezhi Cao", "Jiyuan He", "Yixin Cao"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.21217", "source_url": "https://arxiv.org/abs/2607.21217"}, "categories": ["Agents & Tool Use", "Software & AI Compute", "cs.AI"], "description": {"en": "ICAE-Bench evaluates coding agents on interactive project-building tasks. Agents receive a fuzzy product requirement and must clarify missing details via an automated user agent, then implement the project in a container. Scoring uses black-box tests and multi-dimensional diagnostics including functional correctness, semantic/API similarity, structural fidelity, design quality, and interaction quality."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.21217", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.21217", "title": "ICAE-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Fudan University", "Meituan (China)", "Singapore Management University", "Shanghai Innovation Institute"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.21217", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.21217", "source_url": "https://openalex.org/W7170260613"}, "key": "claire-radar:2607.21217", "modality": null, "name": "ICAE-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_icae-bench_617decda", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "087508d2c4e4cbc072addc9345ef6869f41fa3621e2d68ce0022d2065356dbdb", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.21217"}, "publisher": null, "released": "2026-07-23", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.21217", "source_url": "https://arxiv.org/abs/2607.21217"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-21217", "source": "claire_radar", "source_benchmark_id": "2607.21217", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Software & AI Compute"], "area": "Agents & Tool Use", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/ALEX-nlp/ICAE-EVAL", "githubScope": "benchmark_repo", "githubStars": 7, "hfDailySubmittedAt": "2026-07-23T00:00:00.000Z", "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 8, "hfPaperUrl": "https://huggingface.co/papers/2607.21217", "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": [], "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": "1e44db164316f5d9ccafc513c5c1c94a95245d9fe5bf9c7fe2046f647e0c03fa", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "ICAE-Bench evaluates coding agents on interactive project-building tasks. Agents receive a fuzzy product requirement and must clarify missing details via an automated user agent, then implement the project in a container. Scoring uses black-box tests and multi-dimensional diagnostics including functional correctness, semantic/API similarity, structural fidelity, design quality, and interaction quality.", "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": "In this paper, we introduce ICAE-Bench, a benchmark for evaluating coding agents under interactive project-building settings."}, "familyId": "bmf_0115589cfafb", "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/ALEX-nlp/ICAE-EVAL", "url": "https://github.com/ALEX-nlp/ICAE-EVAL"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/ALEX-nlp/ICAE-EVAL", "url": "https://huggingface.co/papers/2607.21217"}], "scope": "project_repo", "url": "https://github.com/ALEX-nlp/ICAE-EVAL", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_icae-bench_617decda", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Software & Cloud"], "links": {"code": "https://github.com/ALEX-nlp/ICAE-EVAL", "data": null, "hfPaper": "https://huggingface.co/papers/2607.21217", "paper": "https://arxiv.org/abs/2607.21217", "pdf": "https://arxiv.org/pdf/2607.21217", "project": null, "report": "https://arxiv.org/abs/2607.21217"}, "metrics": [], "motivation": "The recent emergence of vibe-coding workflows is changing what coding agents are expected to do.", "name": "ICAE-Bench", "oneLine": "ICAE-Bench evaluates coding agents on interactive project-building tasks. Agents receive a fuzzy product requirement and must clarify missing details via an automated user agent, then implement the project in a container. Scoring uses black-box tests and multi-dimensional diagnostics including functional correctness, semantic/API similarity, structural fidelity, design quality, and interaction quality.", "paperTitle": "ICAE-Bench: Evaluating Coding Agents as Interactive Project Builders", "primaryDomain": "Software & AI Compute", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.21217", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "ALEX-nlp", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/ALEX-nlp/ICAE-EVAL"}], "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": 0.2775680655212807, "value": 7}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": 0.3882377386428991, "value": 8}}, "confidence": "Medium", "coverage": 0.7, "rank": 171, "score": 36}, "method": "attention-ranking-v13", "rank": 171, "score": 36}}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-23", "source": {"authors": ["Zhongyuan Peng", "Dan Huang", "Chuyu Zhang", "Caijun Xu", "Changyi Xiao", "Shibo Hong", "David Lo", "Lin Qiu", "Xuezhi Cao", "Jiyuan He", "Yixin Cao"], "categories": ["cs.AI"], "id": "2607.21217", "title": "ICAE-Bench: Evaluating Coding Agents as Interactive Project Builders", "type": "arxiv", "url": "https://arxiv.org/abs/2607.21217"}, "sourceUpdatedAt": "2026-07-23T11:31:38Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Existing coding benchmarks focus on static, fully specified tasks, leaving a gap for interactive, open-ended development. ICAE-Bench provides a reproducible protocol for measuring agent performance in transforming incomplete requirements into working software, which is increasingly relevant for real-world coding workflows."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
