{"schema_version":"papers/paper-detail-v1","title":"LLMSurgeon: Diagnosing Data Mixture of Large Language Models","surface":"papers","opportunity_kernel":{"paper_id":"2f15c081-0ca9-4120-a591-d00ac736b4ed","title":"LLMSurgeon: Diagnosing Data Mixture of Large Language Models","authors":["Yaxin Luo","Jiacheng Cui","Xiaohan Zhao","Xinyi Shang","Jiacheng Liu","Xinyue Bi","Zhaoyi Li","Zhiqiang Shen"],"arxiv_id":"2605.30348v1","doi":null,"published_at":"2026-05-28T17:59:53.000Z","score_object":{"overall":{"value":7,"scale":"0-10","confidence":0.85,"confidence_reason":"Backfilled from persisted papers.viability_score.","model_version":"phase0-backfill-v1","pipeline_version":"phase0-kernel-v1","computed_at":"2026-05-29T20:12:08.123Z","fresh_until":"2026-06-28T20:12:08.123Z","is_stale":false,"source_count":1,"missingness":[]},"technical":{"value":1.4,"scale":"0-10","confidence":0.55,"confidence_reason":"Backfilled from paper_extraction_scorecards.reconstruction_score.","model_version":"phase0-backfill-v1","pipeline_version":"phase0-kernel-v1","computed_at":"2026-05-29T20:34:18.198Z","fresh_until":"2026-06-12T20:34:18.198Z","is_stale":false,"source_count":1,"missingness":["reproducibility_results.reproducibility_score","deployability_scores.score"]},"commercial":{"value":5,"scale":"0-10","confidence":0.75,"confidence_reason":"Backfilled from persisted commercial_flags and repo availability.","model_version":"phase0-backfill-v1","pipeline_version":"phase0-kernel-v1","computed_at":"2026-05-29T20:12:08.123Z","fresh_until":"2026-06-28T20:12:08.123Z","is_stale":false,"source_count":2,"missingness":[]},"market":{"value":5.9,"scale":"0-10","confidence":0.45,"confidence_reason":"Heuristic market score materialized from paper metadata, repo availability, and deployability.","model_version":"phase0-backfill-v1","pipeline_version":"phase0-kernel-v1","computed_at":"2026-05-29T20:39:51.109Z","fresh_until":"2026-06-12T20:39:51.109Z","is_stale":false,"source_count":1,"missingness":[]},"team":{"value":5.45,"scale":"0-10","confidence":0.42,"confidence_reason":"Heuristic fallback from paper author count and extracted affiliations.","model_version":"phase0-backfill-v1","pipeline_version":"phase0-kernel-v1","computed_at":"2026-05-29T20:12:08.123Z","fresh_until":"2026-06-12T20:12:08.123Z","is_stale":false,"source_count":2,"missingness":[]},"methodology":{"value":4.7,"scale":"0-10","confidence":0.82,"confidence_reason":"Backfilled from paper_extraction_scorecards.total_score.","model_version":"phase0-backfill-v1","pipeline_version":"phase0-kernel-v1","computed_at":"2026-05-29T20:19:07.213Z","fresh_until":"2026-06-28T20:19:07.213Z","is_stale":false,"source_count":2,"missingness":[]}},"evidence_receipt":{"freshness":"fresh","proof_status":"partial","repo_status":"active","references_count":0,"source_count":4,"coverage":0.8333,"missingness":["references"],"unresolved_unknowns":[],"last_verification_at":"2026-05-29T20:34:18.198Z"},"lineage_hash":"e0e4a0a3a9067693c8ccc439e7a28de1c1992d3654a2d1441c2715d8e35c8961"},"distribution":null,"replication_evidence":[],"author_dna":[]}