仅凭模型权重就能验证大模型的传承关系，且无需数据。
Training Leaves Traces: Centered Residual Signatures for Language Model Lineage Verification
- 通过移除共享结构，提取残差块中特有结构差异来判断模型血缘。
- 在多个基准上准确区分微调、剪枝、量化等后代模型，AUROC达1.0。
- 适用于不同模型家族，76倍快于现有方法，适合审计与溯源场景。
开放权重的大语言模型常被微调、量化、剪枝和合并，但其来源往往未记录。我们研究无数据白盒血缘验证：仅凭权重能否判断两个兼容模型检查点是否存在传承关系？残差训练会在分支产物中产生共享的身份对齐成分，因此该结构无法确立血缘。我们移除这一成分，比较残差块间的检查点特异性结构，得到一个相对于独立检查点校准的对称血缘得分。在残差-MLP和GPT-2基准上，该得分能有效区分微调、LoRA合并、剪枝和量化后代模型与独立及蒸馏模型（AUROC=1.0），可区分权重血缘与行为相似性。在保持功能的检查点清洗实验中，传统权重空间基线性能下降或失效；而本方法得分不变，且在GPT-2上比最近的稳健基线快76倍。投影配对信号在六种语言模型族中均存在，并成功识别出3个相关和7个无关的LLaMA-2公开检查点。结果表明，该方法为兼容的开放权重大模型检查点提供了一种被动、无数据的溯源信号。
原文摘要 · Abstract (English)
Open-weight language models are fine-tuned, quantized, pruned, and merged, yet their provenance is often undocumented. We study data-free white-box lineage verification: can weights alone reveal whether two compatible model checkpoints share ancestry? Residual training produces a shared identity-aligned component in branch products, so this structure alone cannot establish ancestry. We remove it and compare checkpoint-specific structure across residual blocks, yielding a symmetric lineage score calibrated against independent checkpoints. On residual-MLP and GPT-2 benchmarks, the score separates fine-tuned, LoRA-merged, pruned, and quantized descendants from independent and distilled models (AUROC=1.0), distinguishing weight ancestry from behavioral similarity. Under function-preserving checkpoint laundering experiments, weight-space baselines lose margin or fail; our score remains unchanged and runs 76x faster than the nearest robust baseline on GPT-2. The projection-pairing signal appears across six language-model families and beyond, and a case study correctly identifies 3 related and 7 unrelated LLaMA-2 public checkpoints. Collectively, these results establish a passive, data-free provenance signal for compatible open-weight language-model checkpoints
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。