Researchers find independently trained transformers compute same function via random rotation

Researchers have discovered a phenomenon called "polymorphism" in independently trained transformers, where they compute the same function but use different internal coordinate systems that are rotated versions of each other. This rotation, which is uniformly random within SO(d_model), makes the internal representations unintelligible between models. However, a single matrix multiplication using an orthogonal Procrustes fit can align these bases, allowing for the transfer of feature dictionaries and steering vectors between models without retraining. AI

影响 Reveals that independently trained models can compute identical functions through rotated internal representations, suggesting potential for cross-model transferability of learned features.

排序理由 The cluster contains an academic paper detailing a new finding about transformer model internals. [lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

• Pythia-70m
• SO(d_model)
• transformer

• 论文
• 模型发布

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →