{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.30248", "kind": "paper", "url": "https://arxiv.org/abs/2608.30248"}], "authors": ["Zhihao Gong", "Junzhe Yu", "Dong Huang", "Zeyu Sun", "Jie M. Zhang", "Dan Hao"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.30248", "source_url": "https://arxiv.org/abs/2608.30248"}, "categories": ["Code", "Code & Software", "Code generation", "Software & AI Compute"], "description": {"en": "Evaluates execution efficiency of LLM-generated data science code across 1,000 tasks using correctness and efficiency scores from stress-testing harnesses."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.30248", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.30248", "title": "DSEffi-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.30248", "modality": null, "name": "DSEffi-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_dseffi-bench_7ed7558d", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "be28e44d186078d4a90a1635a3c486c34c16bcf22b0379dc12e2e07f30817869", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-02T05:23:11.715633Z", "source_url": "https://arxiv.org/abs/2608.30248"}, "publisher": null, "released": "2026-08-31", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.30248", "source_url": "https://arxiv.org/abs/2608.30248"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-30248", "source": "claire_radar", "source_benchmark_id": "2608.30248", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Software & AI Compute"], "area": "Code & Software", "attention": {"asOf": "2026-09-02", "attemptedAt": "2026-09-02T05:12:51.621571Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.30248", "observedAt": "2026-08-31T04:56:28.279469Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-02T05:12:51.621571Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The focus on execution efficiency in code generation addresses a practical gap and includes broad model comparisons, but lacks public artifact links.", "score": 55}, "capabilities": ["Code generation"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-02T05:23:11.715633Z", "inputHash": "199ca8ab273f296d9364e9cfbc5443bd36effe43651609b36bcb984bc012b19a", "model": "deepseek-v4-pro"}, "curation": {"decisionReason": "The named benchmark provides a fixed task set with human-validated references and defined metrics, though no public artifact link is given.", "model": "deepseek-v4-pro", "policy": "source-evidence-v2", "reviewedAt": "2026-09-02T05:23:11.715633Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates execution efficiency of LLM-generated data science code across 1,000 tasks using correctness and efficiency scores from stress-testing harnesses.", "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 DSEffi-Bench, the first benchmark specifically targeting execution efficiency in LLM-generated DS code, comprising 1,000 instances across 10+ DS libraries with stress-testing harnesses and human-validated references."}, "familyId": "bmf_e297347e978c", "firstSeenAt": "2026-09-01", "heat": null, "id": "bm_dseffi-bench_7ed7558d", "indexedAt": "2026-09-02T05:21:06.631990Z", "industrySectors": ["Software & Cloud"], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2608.30248", "pdf": "https://arxiv.org/pdf/2608.30248", "project": null, "report": "https://arxiv.org/abs/2608.30248"}, "metrics": [], "motivation": "Current data science (DS) code generation benchmarks equate correctness with quality, overlooking execution time differences that span orders of magnitude between correct solutions.", "name": "DSEffi-Bench", "oneLine": "Evaluates execution efficiency of LLM-generated data science code across 1,000 tasks using correctness and efficiency scores from stress-testing harnesses.", "paperTitle": "DSEffi-Bench: Demystifying Large Language Models' Capability in Efficient Data Science Code Generation", "primaryDomain": "Software & AI Compute", "publication": {"evidence": "Accepted to EMNLP 2026 Main Conference", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2608.30248", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026 Main Conference", "verifiedAt": "2026-09-02T05:21:06.631990Z"}, "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": "2026-08-31", "source": {"authors": ["Zhihao Gong", "Junzhe Yu", "Dong Huang", "Zeyu Sun", "Jie M. Zhang", "Dan Hao"], "categories": ["cs.SE"], "id": "2608.30248", "title": "DSEffi-Bench: Demystifying Large Language Models' Capability in Efficient Data Science Code Generation", "type": "arxiv", "url": "https://arxiv.org/abs/2608.30248"}, "sourceUpdatedAt": "2026-08-31T04:59:26Z", "topics": ["Code"], "venueAttempts": [{"decisionRaw": "Accepted to EMNLP 2026 Main Conference", "evidence": [{"level": "author-claim", "observedAt": "2026-09-02T05:21:06.631990Z", "rawValue": "Accepted to EMNLP 2026 Main Conference", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2608.30248"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026 Main Conference"}], "whyItMatters": "Shifts data science code evaluation to include efficiency, revealing significant runtime differences and domain-specific failure profiles."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
