Token-Importance Guided Direct Preference Optimization
Ensuring that large language models (LLMs) generate outputs aligned with human preferences is important for safe and effective AI interactions. While Direct Preference Optimization (DPO) employs an implicit reward function to optimize the policy model, however, it and its related variants overlook the differential importance of individual tokens and are sensitive to judgment noise in preference datasets during generation. Although recent methods attempt to assess the important weight of tokens via probability prediction or simplistic weighting schemes, these evaluation methods are prone to biases and still cannot fully address these issues. To solve this problem, we propose the Token-Importance Guided Direct Preference Optimization (TI-DPO), which introduces two key innovations: the gradient-based token-importance weights that dynamically prioritize critical tokens, and a triple loss that explicitly guides model outputs to approach human-preferred responses and stay away from non-preferred responses. Experimental results show that TI-DPO achieves higher accuracy and stronger generative diversity, providing more stable and computationally efficient solutions compared with DPO and other RLHF methods.
Code (0)
등록된 구현이 없습니다.
Tasks
DiversityMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
TIS-DPO: Token-level Importance Sampling for Direct Preference Optimization With Estimated Weights
Direct Preference Optimization (DPO) has been widely adopted for preference alignment of Large Language Models (LLMs) due to its simplicity and effectiveness. However, DPO is derived as a bandit problem in which the whol…
GEM: Generative Entropy-Guided Preference Modeling for Few-shot Alignment of LLMs
Alignment of large language models (LLMs) with human preferences typically relies on supervised reward models or external judges that demand abundant annotations. However, in fields that rely on professional knowledge, s…
Mathematical ReasoningDecision MakingOptimal Transport-Based Token Weighting scheme for Enhanced Preference Optimization
Direct Preference Optimization (DPO) has emerged as a promising framework for aligning Large Language Models (LLMs) with human preferences by directly optimizing the log-likelihood difference between chosen and rejected …
Instruction FollowingToken-weighted Direct Preference Optimization with Attention
Direct Preference Optimization (DPO) aligns Large Language Models with human preferences without the need for a separate reward model. However, DPO treats all tokens in responses equally, neglecting the differing importa…
Towards Bridging the Reward-Generation Gap in Direct Alignment Algorithms
Direct Alignment Algorithms (DAAs), such as Direct Preference Optimization (DPO) and Simple Preference Optimization (SimPO), have emerged as efficient alternatives to Reinforcement Learning from Human Feedback (RLHF) alg…