{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.00384", "kind": "paper", "url": "https://arxiv.org/abs/2609.00384"}, {"id": "gh:mansutti081/restorebench", "kind": "repo", "url": "https://github.com/Mansutti081/RestoreBench"}], "authors": ["Riccardo Mansutti", "Andrea Pomarico", "Robert Jakob", "Qian Zhang", "Alberto Berizzi", "Kevin O'Sullivan"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.00384", "source_url": "https://arxiv.org/abs/2609.00384"}, "categories": ["Agents", "General AI", "Language & Knowledge"], "description": {"en": "RestoreBench evaluates LLM agents on restoring AC power-flow convergence through reactive-control maneuvers in non-convergent power-grid snapshots, with success determined by a physics solver and fixed budget."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.00384", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.00384", "title": "RestoreBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["ETH Zurich", "Politecnico di Milano", "Harvard University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2609.00384", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.00384", "source_url": "https://openalex.org/W7206212011"}, "key": "claire-radar:2609.00384", "modality": null, "name": "RestoreBench", "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": "true", "origin_record_id": "bm_restorebench_236ad72f", "origin_source": "arxiv", "recognition_confidence": "0.95", "record_sha256": "f5c7d157c8b567748f6913bf21c86be5aa22b83978411401aa26c3e0f05d65bf", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-03T00:59:34.317835Z", "source_url": "https://arxiv.org/abs/2609.00384"}, "publisher": null, "released": "2026-08-31", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.00384", "source_url": "https://arxiv.org/abs/2609.00384"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-00384", "source": "claire_radar", "source_benchmark_id": "2609.00384", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-09-02", "attemptedAt": "2026-09-02T05:12:51.621571Z", "githubRepo": "https://github.com/Mansutti081/RestoreBench", "githubScope": "benchmark_repo", "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.00384", "observedAt": "2026-08-31T04:56:28.279469Z", "signalStatus": {"githubStars": {"checkedAt": "2026-09-02T05:12:51.621571Z", "state": "unavailable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-02T05:12:51.621571Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "Specialized domain benchmarks with solver-grounded evaluation tend to attract moderate attention from the AI and power systems communities.", "score": 42}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-03T00:59:34.317835Z", "inputHash": "5285e4b57a7d538e8eb1e9245e18d3fdfabe745924261a3f2a7e7972d5cf55c6", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "**RestoreBench** is a benchmark for **restoring AC power-flow convergence with LLM agents**.", "canonicalNameSource": "official_readme", "decisionReason": "The code repository includes frozen datasets, a standalone scorer, reference agent configurations, and explicit benchmark guarantees, establishing a repeatable evaluation and scoring contract.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-03T00:59:34.317835Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "RestoreBench evaluates LLM agents on restoring AC power-flow convergence through reactive-control maneuvers in non-convergent power-grid snapshots, with success determined by a physics solver and fixed budget.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "We introduce a benchmark that evaluates these capabilities across multiple LLMs and three architectures: \\emph{chatbot}, \\emph{single agent}, and \\emph{multi-agent} systems."}, "familyId": "bmf_88bcdfe14566", "firstSeenAt": "2026-09-02", "heat": null, "id": "bm_restorebench_236ad72f", "indexedAt": "2026-09-02T00:47:12.074634Z", "industrySectors": [], "links": {"code": "https://github.com/Mansutti081/RestoreBench", "data": null, "paper": "https://arxiv.org/abs/2609.00384", "pdf": "https://arxiv.org/pdf/2609.00384", "project": null, "report": "https://arxiv.org/abs/2609.00384"}, "metrics": [], "motivation": "Large Language Model (LLM) agents increasingly automate multi-step engineering workflows through tool use, interpretation of intermediate results, and iterative planning.", "name": "RestoreBench", "oneLine": "RestoreBench evaluates LLM agents on restoring AC power-flow convergence through reactive-control maneuvers in non-convergent power-grid snapshots, with success determined by a physics solver and fixed budget.", "paperTitle": "RestoreBench: Can AI Agents Restore Power Flow Convergence?", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.00384", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-02T00:47:12.074634Z"}, "publications": [], "publishers": [{"name": "Mansutti081", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/Mansutti081/RestoreBench"}], "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": "Runnable", "recognitionConfidence": 0.95, "relation": "introduces", "releasedAt": "2026-08-31", "source": {"authors": ["Riccardo Mansutti", "Andrea Pomarico", "Robert Jakob", "Qian Zhang", "Alberto Berizzi", "Kevin O'Sullivan"], "categories": ["cs.AI", "cs.SY", "eess.SY"], "id": "2609.00384", "title": "RestoreBench: Can AI Agents Restore Power Flow Convergence?", "type": "arxiv", "url": "https://arxiv.org/abs/2609.00384"}, "sourceUpdatedAt": "2026-08-31T21:15:19Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "It provides a solver-grounded testbed for agent reasoning in power systems, where each action is verified by pandapower, exposing ungrounded reasoning in a safety-critical domain."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
