{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.04200", "kind": "paper", "url": "https://arxiv.org/abs/2609.04200"}, {"kind": "website", "url": "https://principiabench.github.io/"}], "authors": ["Varun Varma Thozhiyoor", "Shivam Tripathi", "Venkatesh Babu Radhakrishnan", "Anand Bhattad"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.04200", "source_url": "https://arxiv.org/abs/2609.04200"}, "categories": ["General AI", "Vision & 3D", "cs.CV"], "description": {"en": "Evaluates Newtonian physics reasoning in video models through relational consistency between paired objects across eight phenomena, using a calibration-independent consistency score derived from real-world scenes."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.04200", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.04200", "title": "Principia"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.04200", "modality": null, "name": "Principia", "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_principia_eb88a792", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "cb547de8516e2aa3065636b9b9c4d3823a27043cd7425767a26e410cd9adefc2", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-07T00:42:15.394147Z", "source_url": "https://arxiv.org/abs/2609.04200"}, "publisher": null, "released": "2026-09-03", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.04200", "source_url": "https://arxiv.org/abs/2609.04200"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-04200", "source": "claire_radar", "source_benchmark_id": "2609.04200", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-09-04", "attemptedAt": "2026-09-04T06:36:46.783164Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": "2026-09-04T00:00:00.000Z", "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 2, "hfPaperUrl": "https://huggingface.co/papers/2609.04200", "observedAt": "2026-09-04T06:36:46.783164Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-09-04T06:36:46.783164Z", "state": "fresh"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "Physics reasoning in video models is a high-interest topic and the benchmark offers a novel metric with a project page, likely attracting attention from the video generation and evaluation communities.", "score": 62}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-07T00:42:15.394147Z", "inputHash": "0381b4d8391542b517b134c6aecf396025789cab295a21814eceb16fac444f05", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "We introduce Principia, a benchmark that evaluates Newtonian physics through relational consistency between paired objects.", "canonicalNameSource": "abstract", "decisionReason": "The benchmark has an official project page and a defined scoring metric, though no code or data is directly linked; the project page provides a public reference point.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-07T00:42:15.394147Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates Newtonian physics reasoning in video models through relational consistency between paired objects across eight phenomena, using a calibration-independent consistency score derived from real-world scenes.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "We introduce Principia, a benchmark that evaluates Newtonian physics through relational consistency between paired objects."}, "familyId": "bmf_03c87d9a6cfc", "firstSeenAt": "2026-09-04", "heat": null, "id": "bm_principia_eb88a792", "indexedAt": "2026-09-04T06:13:54.443310Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.04200", "pdf": "https://arxiv.org/pdf/2609.04200", "project": "https://principiabench.github.io/", "report": "https://arxiv.org/abs/2609.04200"}, "metrics": [], "motivation": "Evaluating physical reasoning in video models is difficult because absolute motion measurements depend on frame rate, object scale, and camera calibration, all of which are often ambiguous or unavailable in generated video.", "name": "Principia", "oneLine": "Evaluates Newtonian physics reasoning in video models through relational consistency between paired objects across eight phenomena, using a calibration-independent consistency score derived from real-world scenes.", "paperTitle": "Principia: Relational Physics Tests for Video Models", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.04200", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-04T06:13:54.443310Z"}, "publications": [], "publishers": [{"name": "Principia Bench", "organizationType": "benchmark-organization", "role": "benchmark-publisher", "sourceUrl": "https://principiabench.github.io/"}], "ranking": {"30d": {"confidence": "Low", "coverage": 0.3, "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": 0.19460147417678111, "value": 2}}, "confidence": "Low", "coverage": 0.3, "rank": null, "score": 41}, "method": "attention-ranking-v13", "rank": null, "score": 41}, "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": 0.19460147417678111, "value": 2}}, "confidence": "Low", "coverage": 0.15, "rank": null, "score": 45}, "method": "attention-ranking-v13", "rank": null, "score": 45}, "today": {"confidence": "Low", "coverage": 0.45, "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": 1, "historyDate": "2026-09-03", "rank": null, "score": null, "windowDays": 1}, "level": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": 0.5, "percentile": 0.5, "value": 2}, "llmAttentionForecast": {"maxContribution": 0.25, "percentile": 0.7, "role": "bonus", "value": 70}}, "confidence": "Low", "coverage": 0.45, "rank": 2, "score": 55}, "method": "attention-ranking-v13", "rank": 2, "score": 55}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-03", "source": {"authors": ["Varun Varma Thozhiyoor", "Shivam Tripathi", "Venkatesh Babu Radhakrishnan", "Anand Bhattad"], "categories": ["cs.CV"], "id": "2609.04200", "title": "Principia: Relational Physics Tests for Video Models", "type": "arxiv", "url": "https://arxiv.org/abs/2609.04200"}, "sourceUpdatedAt": "2026-09-03T17:59:50Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "Provides a calibration-independent measure of physical plausibility in generated video, revealing that current models perform far worse than standard video quality benchmarks suggest."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
