← 研踪 PaperTrailarXiv:2309.00941 · v2
Emergent Linear Representations in World Models of Self-Supervised Sequence Models
首次提交 2023-09-02 21:37
版本更新 2023-09-08 04:36
雷达观测 2026-10-29 15:13
摘要
How do sequence models represent their decision-making process? Prior work suggests that Othello-playing neural network learned nonlinear models of the board state (Li et al., 2023). In this work, we provide evidence of a closely related linear representation of the board. In particular, we show that probing for "my colour" vs. "opponent's colour" may be a simple yet powerful way to interpret the model's internal state. This precise understanding of the internal representations allows us to control the model's behaviour with simple vector arithmetic. Linear representations enable significant interpretability progress, which we demonstrate with further exploration of how the world model is computed.
论文发现出处
这些页面提供论文线索，不代表同行评审或质量认证。
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v2 · 2023-09-08 04:36 · 最新观测版本
Emergent Linear Representations in World Models of Self-Supervised Sequence Models
How do sequence models represent their decision-making process? Prior work suggests that Othello-playing neural network learned nonlinear models of the board state (Li et al., 2023). In this work, we provide evidence of a closely related linear representation of the board. In particular, we show that probing for "my colour" vs. "opponent's colour" may be a simple yet powerful way to interpret the model's internal state. This precise understanding of the internal representations allows us to control the model's behaviour with simple vector arithmetic. Linear representations enable significant interpretability progress, which we demonstrate with further exploration of how the world model is computed.
查看此版本解读 →查看此版本原文 ↗