Policy Iteration with Human Feedback: Bringing Post-Training RL to In-context Learning
Generative pretraining established reusable task representations; later work on language-based task conditioning and in-context learning showed that a fixed model could adapt its behavior from instructions and demonstrations. Policy Iteration with Human Feedback (PIHF) builds on this development and the recurrent evaluate-and-improve structure of generalized policy iteration. PIHF uses a pretrained language model as its execution substrate and moves persistent revision to a versioned natural-language policy and tool set. A language-model critic and clinical expert review complete-panel reasoning and tool-use trajectories to localize recurrent failures and form candidate revisions; the expert may reinterpret the evidence and retains authority over admission and rollback, while Recall@1 and Recall@5 validate outcomes after candidate execution. Across cumulative ablations and ultra-rare-disease benchmarks, a PIHF-derived policy improved Recall@1 in one proprietary executor and three open-weight executors spanning 3 to 49 billion active parameters. Gains were 32.7 percentage points for GPT-5.4 and 31.1 points for Qwen3.6-35B, a difference of 1.7 points. These results support the feasibility of using pretrained language models as fixed-weight execution substrates for expert-guided policy development in rare-disease diagnosis.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Policy Improvement Reinforcement Learning
Reinforcement learning has become a central post-training paradigm for improving LLM and agent capabilities. Yet existing RL post-training methods share a common blind spot: they construct local learning signals from sam…
Mathematical ReasoningReinforcement LearningCooperative Optimal Output Tracking for Discrete-Time Multiagent Systems: Stabilizing Policy Iteration Frameworks and Analysis
In this paper, two model-free optimal output tracking frameworks based on policy iteration for discrete-time multi-agent systems are proposed. First, we establish a framework of stabilizing policy iteration that can star…
Q-LearningZeroth-Order Policy Gradient for Reinforcement Learning from Human Feedback without Reward Inference
Reward inference (learning a reward model from human preferences) is a critical intermediate step in Reinforcement Learning from Human Feedback (RLHF) for fine-tuning Large Language Models (LLMs) such as ChatGPT. In prac…
Reinforcement Learning with Human Feedback: Learning Dynamic Choices via Pessimism
In this paper, we study offline Reinforcement Learning with Human Feedback (RLHF) where we aim to learn the human's underlying reward and the MDP's optimal policy from a set of trajectories induced by human choices. RLHF…
Decision MakingEconometricsOffline RLreinforcement-learning+1Policy Gradient Primal-Dual Method for Safe Reinforcement Learning from Human Feedback
Safe Reinforcement Learning from Human Feedback (Safe RLHF) has recently achieved empirical success in developing helpful and harmless large language models by decoupling human preferences regarding helpfulness and harml…
Reinforcement Learning