{"version":1,"type":"story","url":"https://digestai.news/story/researchers-propose-witness-overlap-to-track-open-weight-model-lineage","json":"https://digestai.news/story/researchers-propose-witness-overlap-to-track-open-weight-model-lineage.json","markdown":"https://digestai.news/story/researchers-propose-witness-overlap-to-track-open-weight-model-lineage.md","slug":"researchers-propose-witness-overlap-to-track-open-weight-model-lineage","headline":"Researchers propose Witness Overlap to track Open-Weight model lineage","summary":"Witeness Overlap: Directional Provenance Inside Open-Weight Model Families\narXiv:2609.31784v1 Announce Type: new \nAbstract: Open-weight models are often released, fine-tuned, aligned, merged, and re-released, making provenance audits ask not only whether checkpoints are related, but also which checkpoint came first. Many existing model-provenance methods are designed for a base-known audit setting: given a victim or source model, they test whether a suspect model is related to it. Although these audits are framed as source-to-suspect tests, their underlying evidence is often symmetric, relying on representation similarity, weight similarity, behavioral fingerprints, or correlation statistics. Symmetric pairwise comparisons can detect relatedness, but they cannot by themselves orient relationship between checkpoints A and B. We therefore introduce a local geometric comparison: instead of comparing two checkpoints directly, we add a third same-family checkpoint as a witness and compare the geometry around each candidate endpoint.","keyPoints":["Witness Overlap uses a third model as a reference to infer directional relationships between two checkpoints","Method achieves 95.3% accuracy in orienting parent-child decisions across 176 LLM checkpoints from 16 families","Test works without training data or prompts and remains robust to weight noise and pruning"],"whyItMatters":"This method could help researchers and developers trace lineage and detect unauthorized model derivations in open-weight AI models.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":[],"models":[],"people":[]},"firstPublishedAt":"2026-09-29T04:00:00Z","updatedAt":"2026-09-29T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"Witeness Overlap: Directional Provenance Inside Open-Weight Model Families","url":"https://arxiv.org/abs/2609.31784","publishedAt":"2026-09-29T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Researchers propose Witness Overlap to track Open-Weight model lineage\", 29 September 2026, https://digestai.news/story/researchers-propose-witness-overlap-to-track-open-weight-model-lineage","publisher":"Digest AI","title":"Researchers propose Witness Overlap to track Open-Weight model lineage","datePublished":"2026-09-29T04:00:00Z","url":"https://digestai.news/story/researchers-propose-witness-overlap-to-track-open-weight-model-lineage"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}