发现多模态指令微调让模型用大小而非方向编码身份信息
From Direction to Magnitude: How Multimodal Instruction-Tuning Reorganizes the Geometric Encoding of Identity-Specifying Prompts in Transformer Hidden States
- 通过曲率分布距离分析隐藏层轨迹，揭示身份提示的几何指纹变化
- 指令微调后身份编码从方向转为幅度，基线模型方向差异显著（p=0.002）
- 该现象仅在多模态指令微调中出现，适合研究模型表征机制的学者
我们研究了身份指定型系统提示是否在四种开放权重Transformer语言模型的隐藏状态轨迹中产生可区分的几何指纹，这些模型涵盖四种训练阶段：无训练（Gemma-4-E4B base）、多模态RLHF（Gemma-4-E4B-it）、RL蒸馏（DeepSeek-R1-Distill-Qwen-7B）和SFT（Qwen2.5-7B-Instruct）。比较三种提示条件（身份轴提示、长度匹配的通用助手提示、26令牌基线），使用五种几何度量，核心是基于k-NN轨迹图上边际分布的Ollivier-Ricci曲率的1-Wasserstein距离。分析基于轨迹级置换检验，包含多重几何控制（教师强制内容控制、时序链与k-NN拓扑对比、ABT投影k-NN、角度与欧氏图构建、5000次置换以验证边界统计量）。核心发现为：在指令微调边界处，身份编码发生定性重构——基线模型中指纹以方向编码为主（角度分离0.034，p=0.002，角度k-NN下）；多模态指令微调模型中其迁移至幅度（角度分离降至p=0.439，欧氏仍显著，p=0.042，首个生成状态的平均范数反转长度排序，身份提示最低）。此方向到幅度的重构仅出现在多模态指令微调，其余训练方式中未见。教师强制控制表明，约30%的自由运行余弦信号来自提示驱动效应。我们提出将W_1应用于边际曲率分布作为独立方法贡献。
原文摘要 · Abstract (English)
We investigate whether identity-specifying system prompts produce statistically distinguishable geometric fingerprints in the hidden-state trajectories of four open-weight transformer language models spanning four post-training regimes: no training (Gemma-4-E4B base), multimodal RLHF (Gemma-4-E4B-it), RL distillation (DeepSeek-R1-Distill-Qwen-7B), and SFT (Qwen2.5-7B-Instruct). Three prompt conditions (an identity-specifying axis prompt, a length-matched generic-assistant prompt, and a 26-token vanilla baseline) are compared via five geometric metrics, principally the 1-Wasserstein distance between edge-wise distributions of Ollivier-Ricci curvature on k-NN trajectory graphs. Claims rest on trajectory-level permutation tests with multiple geometric controls (teacher-forced content controls, temporal-chain vs k-NN topology, ABT-projected k-NN, angular vs Euclidean graph construction, B=5000 permutations on borderline statistics). The central finding is a qualitative reorganization of identity encoding across the instruction-tuning boundary: in the base model the fingerprint is direction-coded (separation 0.034, p=0.002 under angular k-NN); in the multimodal instruction-tuned model it migrates into the magnitude (angular separation collapses to p=0.439 while Euclidean survives at p=0.042, and the mean norm of the first generated state inverts its length-ordering, being lowest for the identity prompt). This direction-to-magnitude reorganization is specific to the multimodal instruction-tuning regime, absent under RL distillation and SFT. A teacher-forced control attributes ~30% of the free-running cosine signal to prompt-driven effects. We position W_1 on edge-wise Ollivier-Ricci distributions on k-NN trajectory graphs as a methodological contribution of independent interest.
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。