{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.18789", "kind": "paper", "url": "https://arxiv.org/abs/2607.18789"}], "authors": ["Amber Yijia Zheng", "Lu Liu", "Raymond A. Yeh", "Xi Yin"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.18789", "source_url": "https://arxiv.org/abs/2607.18789"}, "categories": ["General AI", "Vision & 3D", "cs.CV"], "description": {"en": "Moving Alphabet is a procedural testbed for controlled experiments on text-to-video training data, generating synthetic videos with ground-truth metadata to study data distribution and caption quality effects."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.18789", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.18789", "title": "Moving Alphabet"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Purdue University West Lafayette"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2607.18789", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2607.18789", "source_url": "https://openalex.org/W7170046785"}, "key": "claire-radar:2607.18789", "modality": null, "name": "Moving Alphabet", "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_moving-alphabet_5bd200ea", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "e0938221d157173e414be5c2d12cffa594bbcff28487d7b0b38fd7a291a3f619", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.18789"}, "publisher": null, "released": "2026-07-21", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.18789", "source_url": "https://arxiv.org/abs/2607.18789"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-18789", "source": "claire_radar", "source_benchmark_id": "2607.18789", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": "2026-07-23T00:00:00.000Z", "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 3, "hfPaperUrl": "https://huggingface.co/papers/2607.18789", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "capabilities": [], "confidence": "High", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "2354ee3f9f8ea9b6123774383feef61664bc9ab43e70622b142ddf0dbd5c59c8", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Moving Alphabet is a procedural testbed for controlled experiments on text-to-video training data, generating synthetic videos with ground-truth metadata to study data distribution and caption quality effects.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence"], "snippet": "To enable controlled experiments, we introduce Moving Alphabet, a procedural testbed that renders letters with varying fonts, colors, sizes, and positions, moving in different directions and speeds against a black background."}, "familyId": "bmf_9c1339291416", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_moving-alphabet_5bd200ea", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2607.18789", "paper": "https://arxiv.org/abs/2607.18789", "pdf": "https://arxiv.org/pdf/2607.18789", "project": null, "report": "https://arxiv.org/abs/2607.18789"}, "metrics": [], "motivation": "Text-to-video generation has advanced significantly over the past five years through scaling of model size, data, and compute.", "name": "Moving Alphabet", "oneLine": "Moving Alphabet is a procedural testbed for controlled experiments on text-to-video training data, generating synthetic videos with ground-truth metadata to study data distribution and caption quality effects.", "paperTitle": "Moving Alphabet: A Controlled Study of Training Data for Text-to-Video Generation", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.18789", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "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": 0.24495074076910422, "value": 3}}, "confidence": "Low", "coverage": 0.15, "rank": null, "score": 46}, "method": "attention-ranking-v13", "rank": null, "score": 46}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-21", "source": {"authors": ["Amber Yijia Zheng", "Lu Liu", "Raymond A. Yeh", "Xi Yin"], "categories": ["cs.CV"], "id": "2607.18789", "title": "Moving Alphabet: A Controlled Study of Training Data for Text-to-Video Generation", "type": "arxiv", "url": "https://arxiv.org/abs/2607.18789"}, "sourceUpdatedAt": "2026-07-21T07:07:19Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "It provides insights into data curation for text-to-video models, highlighting the importance of diverse distributions and caption quality, but it is not a model comparison benchmark."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
