Controlled Decoding from Language Models
KL-regularized reinforcement learning (RL) is a popular alignment framework to control the language model responses towards high reward outcomes. We pose a tokenwise RL objective and propose a modular solver for it, called controlled decoding (CD). CD exerts control through a separate prefix scorer module, which is trained to learn a value function for the reward. The prefix scorer is used at inference time to control the generation from a frozen base model, provably sampling from a solution to the RL objective. We empirically demonstrate that CD is effective as a control mechanism on popular benchmarks. We also show that prefix scorers for multiple rewards may be combined at inference time, effectively solving a multi-objective RL problem with no additional training. We show that the benefits of applying CD transfer to an unseen base model with no further tuning as well. Finally, we show that CD can be applied in a blockwise decoding fashion at inference-time, essentially bridging the gap between the popular best-of-K strategy and tokenwise control through reinforcement learning. This makes CD a promising approach for alignment of language models.
Code (0)
등록된 구현이 없습니다.
Tasks
Language ModelingLanguage ModellingMulti-Objective Reinforcement Learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Critic-Guided Decoding for Controlled Text Generation
Steering language generation towards objectives or away from undesired content has been a long-standing goal in utilizing language models (LM). Recent work has demonstrated reinforcement learning and weighted decoding as…
Language ModelingLanguage Modellingreinforcement-learningReinforcement Learning+2DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts
Despite recent advances in natural language generation, it remains challenging to control attributes of generated text. We propose DExperts: Decoding-time Experts, a decoding-time method for controlled text generation th…
Language ModelingLanguage ModellingText GenerationControlled LLM Decoding via Discrete Auto-regressive Biasing
Controlled text generation allows for enforcing user-defined constraints on large language model outputs, an increasingly important field as LLMs become more prevalent in everyday life. One common approach uses energy-ba…
Large Language ModelText GenerationVisual-Redundancy-Controlled Parallel Decoding for Diffusion-Based Multimodal Large Language Models
Diffusion-based multimodal large language models (dMLLMs) decode by iteratively predicting tokens at multiple masked positions in parallel. This turns each decoding step into a position-selection problem: the model must …
Visual GroundingGrammar Control in Dialogue Response Generation for Language Learning Chatbots
Chatbots based on large language models offer cheap conversation practice opportunities for language learners. However, they are hard to control for linguistic forms that correspond to learners' current needs, such as gr…
ChatbotLanguage AcquisitionResponse Generation