# Witness Overlap: Directional Provenance Inside Open-Weight Model Families

arXiv: https://arxiv.org/abs/2609.31784 · HTML: https://mlsanity.com/paper/2609.31784/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 21 references that are in the map.
- First posted on arXiv 24 Sep 2026.

## Builds on (most cited first, 16 of 16)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [PubMedQA: A Dataset for Biomedical Research Question Answering](https://mlsanity.com/paper/1909.06146/) (2019)
- [Model Cards for Model Reporting](https://mlsanity.com/paper/1810.03993/) (2018)
- [Enhancing Chat Language Models by Scaling High-quality Instructional Conversations](https://mlsanity.com/paper/2305.14233/) (2023)
- [TIES-Merging: Resolving Interference When Merging Models](https://mlsanity.com/paper/2306.01708/) (2023)
- [An Empirical Study of Pre-Trained Model Reuse in the Hugging Face Deep Learning Model Registry](https://mlsanity.com/paper/2303.02552/) (2023)
- [REEF: Representation Encoding Fingerprints for Large Language Models](https://mlsanity.com/paper/2410.14273/) (2024)
- [Model Provenance Testing for Large Language Models](https://mlsanity.com/paper/2502.00706/) (2025)
- [LLM DNA: Tracing Model Evolution via Functional Representations](https://mlsanity.com/paper/2509.24496/) (2025)
- [Unsupervised Model Tree Heritage Recovery](https://mlsanity.com/paper/2405.18432/) (2024)
- [Blackbox Model Provenance via Palimpsestic Membership Inference](https://mlsanity.com/paper/2510.19796/) (2025)
- [Independence Tests for Language Models](https://mlsanity.com/paper/2502.12292/) (2025)
- [Neural Lineage](https://mlsanity.com/paper/2406.11129/) (2024)
- [HuggingGraph: Understanding the Supply Chain of LLM Ecosystem](https://mlsanity.com/paper/2507.14240/) (2025)
- [An Empirical Analysis of Machine Learning Model and Dataset Documentation, Supply Chain, and Licensing Challenges on Hugging Face](https://mlsanity.com/paper/2502.04484/) (2025)
- [AttnDiff: Attention-based Differential Fingerprinting for Large Language Models](https://mlsanity.com/paper/2604.05502/) (2026)

## Abstract (arXiv)

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. Direction is inferred by asking which candidate behaves more like a branching parent. Motivated by this idea, and by the empirically observed asymmetry between parent-anchored and child-anchored witness-overlap distributions, we propose Witness Overlap, a prompt-free, training-free white-box test for directional provenance. On 176 LLM checkpoints from 16 families, our one-witness test orients 95.3% of parent-child decisions using Frobenius cosine. We further evaluate root identification, sibling discrimination, generalizations to VLM and diffusion families, and chain-structured ordering. The signal is robust to weight noise and sparse pruning, with a proposed SVD weight reduction variant showing greater robustness than Frobenius cosine.