Constrain Alignment with Sparse Autoencoders
This addresses the problem of computational inefficiency and instability in LLM alignment for AI developers, representing an incremental improvement over existing techniques.
The paper tackles the challenge of aligning large language models with human preferences by proposing Feature-level constrained Preference Optimization (FPO), which uses sparse autoencoders and feature-level constraints to achieve a 5.08% absolute improvement in win rate with lower computational cost compared to state-of-the-art methods.
The alignment of large language models (LLMs) with human preferences remains a key challenge. While post-training techniques like Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) have achieved notable success, they often introduce computational inefficiencies and training instability. In this paper, we propose Feature-level constrained Preference Optimization (FPO), a novel method designed to simplify the alignment process while ensuring stability. FPO leverages pre-trained Sparse Autoencoders (SAEs) and introduces feature-level constraints, allowing for efficient, sparsity-enforced alignment. Our approach enjoys efficiency by using sparse features activated in a well-trained sparse autoencoder and the quality of sequential KL divergence by using the feature-level offline reference. Experimental results on benchmark datasets demonstrate that FPO achieves a 5.08% absolute improvement in win rate with much lower computational cost compared to state-of-the-art baselines, making it a promising solution for efficient and controllable LLM alignments.