{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.30241", "kind": "paper", "url": "https://arxiv.org/abs/2608.30241"}, {"kind": "website", "url": "https://shirley-wu.github.io/PaperBanana-Interact/"}], "categories": ["General AI", "Language & Knowledge", "cs.CL"], "description": {"en": "Evaluates multi-turn scientific diagram refinement via 292 images annotated with 3,518 user requirements, using a user simulator that provides iterative feedback and measures requirement satisfaction and diagram quality."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.30241", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.30241", "title": "MTPaperBananaBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.30241", "modality": null, "name": "MTPaperBananaBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "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_mtpaperbananabench_c4c3a34a", "origin_source": "arxiv", "recognition_confidence": "0.55", "record_sha256": "f82796e5f206e9c20a9913a85f8f19f397b6a40320faff6aacfd65d7bdb57e3b", "relation": "evaluates_only", "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/2608.30241"}, "publisher": null, "released": "2026-08-31", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.30241", "source_url": "https://arxiv.org/abs/2608.30241"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-30241", "source": "claire_radar", "source_benchmark_id": "2608.30241", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 11, "hfPaperUrl": "https://huggingface.co/papers/2608.30241", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-10-01", "state": "stale"}}, "status": "stale"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The benchmark targets a niche multi-turn scientific diagram task with moderate data size, likely drawing interest from diagram generation and agentic refinement communities but lacking broad appeal.", "score": 55}, "capabilities": [], "confidence": "Medium", "construction": "Interactive Environment", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-02T05:23:11.715633Z", "inputHash": "1c96d2f6c1a317142d9152241acb1eb21a057e8d12cdf45a32e4dadf6e765deb", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "we present MTPaperBananaBench, a benchmark for multi-turn diagram generation containing 292 images annotated with 3,518 user requirements.", "canonicalNameSource": "abstract", "decisionReason": "The paper defines a repeatable benchmark with annotated data and a user simulator, establishes scoring via requirement satisfaction and quality, though the lack of a public data or code link lowers readiness but does not preclude the benchmark's formal release.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-02T05:23:11.715633Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates multi-turn scientific diagram refinement via 292 images annotated with 3,518 user requirements, using a user simulator that provides iterative feedback and measures requirement satisfaction and diagram quality.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL", "named entity is a model, method, framework, or dataset used in evaluation"], "snippet": "To bridge this gap, we present MTPaperBananaBench, a benchmark for multi-turn diagram generation containing 292 images annotated with 3,518 user requirements."}, "familyId": "bmf_753da8a51c06", "firstSeenAt": "2026-09-01", "heat": null, "id": "bm_mtpaperbananabench_c4c3a34a", "indexedAt": "2026-09-01T06:28:29.895112Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2608.30241", "pdf": "https://arxiv.org/pdf/2608.30241", "project": "https://shirley-wu.github.io/PaperBanana-Interact/", "report": "https://arxiv.org/abs/2608.30241"}, "metrics": [], "motivation": "Recent efforts have aimed to automate scientific diagram generation from paper content (Lin et al., 2026; Zhu et al., 2026a).", "name": "MTPaperBananaBench", "oneLine": "Evaluates multi-turn scientific diagram refinement via 292 images annotated with 3,518 user requirements, using a user simulator that provides iterative feedback and measures requirement satisfaction and diagram quality.", "paperTitle": "PaperBanana-Interact: Scientific Diagram Refinement with Multi-Turn Human Feedback", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.30241", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-01T06:28:29.895112Z"}, "publications": [], "ranking": {"90d": {"confidence": "Low", "coverage": 0.15, "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": 0.4390696100905538, "value": 11}}, "confidence": "Low", "coverage": 0.15, "rank": null, "score": 49}, "method": "attention-ranking-v13", "rank": null, "score": 49}}, "readiness": "Paper only", "recognitionConfidence": 0.55, "relation": "evaluates_only", "releasedAt": "2026-08-31", "source": {"authors": ["Xueqing Wu", "Ashwin Balasubramanian", "Bingxuan Li", "Dawei Zhu", "Kai-Wei Chang", "Yale Song", "Yiwen Song", "Rui Meng", "Tomas Pfister", "Nanyun Peng"], "categories": ["cs.CL", "cs.CV"], "id": "2608.30241", "title": "PaperBanana-Interact: Scientific Diagram Refinement with Multi-Turn Human Feedback", "type": "arxiv", "url": "https://arxiv.org/abs/2608.30241"}, "sourceUpdatedAt": "2026-08-31T04:52:32Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Addresses the underexplored multi-turn diagram generation workflow, revealing failure modes like quality drift and forgetting that single-turn benchmarks miss, and provides a scalable way to test iterative refinement systems."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
