Optimal Policy Minimum Bayesian Risk

• URL: http://arxiv.org/abs/2505.17242v1
• Date: Thu, 22 May 2025 19:43:37 GMT
• Title: Optimal Policy Minimum Bayesian Risk
• Authors: Ramón Fernandez Astudillo, Md Arafat Sultan, Aashka Trivedi, Yousef El-Kurdi, Tahira Naseem, Radu Florian, Salim Roukos,
• Abstract summary: We present a novel method for incorporating reward and risk/similarity signals into Bayes risk decoding (MBRD)<n>Based on the concept of optimal policy in KL-controlled reinforcement learning, our framework provides a simple and well-defined mechanism for leveraging such signals.<n>It offers several advantages over traditional inference-time methods: higher robustness, improved accuracy, and well-understood behavior.
• Score: 25.434911234706952
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Inference scaling can help LLMs solve complex reasoning problems through extended runtime computation. On top of targeted supervision for long chain-of-thought (long-CoT) generation, purely inference-time techniques such as best-of-N (BoN) sampling, majority voting, or more generally, minimum Bayes risk decoding (MBRD), can further improve LLM accuracy by generating multiple candidate solutions and aggregating over them. These methods typically leverage additional signals in the form of reward models and risk/similarity functions that compare generated samples, e.g., exact match in some normalized space or standard similarity metrics such as Rouge. Here we present a novel method for incorporating reward and risk/similarity signals into MBRD. Based on the concept of optimal policy in KL-controlled reinforcement learning, our framework provides a simple and well-defined mechanism for leveraging such signals, offering several advantages over traditional inference-time methods: higher robustness, improved accuracy, and well-understood asymptotic behavior. In addition, it allows for the development of a sample-efficient variant of MBRD that can adjust the number of samples to generate according to the difficulty of the problem, without relying on majority vote counts. We empirically demonstrate the advantages of our approach on math (MATH-$500$) and coding (HumanEval) tasks using recent open-source models. We also present a comprehensive analysis of its accuracy-compute trade-offs.

Related papers

• Collab: Controlled Decoding using Mixture of Agents for LLM Alignment [90.6117569025754]
  Reinforcement learning from human feedback has emerged as an effective technique to align Large Language models.<n>Controlled Decoding provides a mechanism for aligning a model at inference time without retraining.<n>We propose a mixture of agent-based decoding strategies leveraging the existing off-the-shelf aligned LLM policies.
  arXiv  Detail & Related papers  (2025-03-27T17:34:25Z)
• Adaptive Sampled Softmax with Inverted Multi-Index: Methods, Theory and Applications [79.53938312089308]
  The MIDX-Sampler is a novel adaptive sampling strategy based on an inverted multi-index approach.<n>Our method is backed by rigorous theoretical analysis, addressing key concerns such as sampling bias, gradient bias, convergence rates, and generalization error bounds.
  arXiv  Detail & Related papers  (2025-01-15T04:09:21Z)
• Burning RED: Unlocking Subtask-Driven Reinforcement Learning and Risk-Awareness in Average-Reward Markov Decision Processes [7.028778922533688]
  Average-reward Markov decision processes (MDPs) provide a foundational framework for sequential decision-making under uncertainty.<n>We introduce Reward-Extended Differential (or RED) reinforcement learning: a novel RL framework that can be used to effectively and efficiently solve various learning objectives, or subtasks, simultaneously in the average-reward setting.
  arXiv  Detail & Related papers  (2024-10-14T14:52:23Z)
• Gaussian Stochastic Weight Averaging for Bayesian Low-Rank Adaptation of Large Language Models [5.352221132808875]
  Fine-tuned Large Language Models (LLMs) often suffer from overconfidence and poor calibration. We propose a simple combination of Low-Rank Adaptation (LoRA) with Gaussian Weight Averaging (SWAG) We show that our method exhibits greater robustness against distribution shift, as reflected in its improved performance on out-of-distribution tasks.
  arXiv  Detail & Related papers  (2024-05-06T12:44:37Z)
• Provably Efficient Information-Directed Sampling Algorithms for Multi-Agent Reinforcement Learning [50.92957910121088]
  This work designs and analyzes a novel set of algorithms for multi-agent reinforcement learning (MARL) based on the principle of information-directed sampling (IDS) For episodic two-player zero-sum MGs, we present three sample-efficient algorithms for learning Nash equilibrium. We extend Reg-MAIDS to multi-player general-sum MGs and prove that it can learn either the Nash equilibrium or coarse correlated equilibrium in a sample efficient manner.
  arXiv  Detail & Related papers  (2024-04-30T06:48:56Z)
• Best Arm Identification with Fixed Budget: A Large Deviation Perspective [54.305323903582845]
  We present sred, a truly adaptive algorithm that can reject arms in it any round based on the observed empirical gaps between the rewards of various arms. In particular, we present sred, a truly adaptive algorithm that can reject arms in it any round based on the observed empirical gaps between the rewards of various arms.
  arXiv  Detail & Related papers  (2023-12-19T13:17:43Z)
• Distributionally Robust Models with Parametric Likelihood Ratios [123.05074253513935]
  Three simple ideas allow us to train models with DRO using a broader class of parametric likelihood ratios. We find that models trained with the resulting parametric adversaries are consistently more robust to subpopulation shifts when compared to other DRO approaches.
  arXiv  Detail & Related papers  (2022-04-13T12:43:12Z)
• Adaptive Sampling for Best Policy Identification in Markov Decision Processes [79.4957965474334]
  We investigate the problem of best-policy identification in discounted Markov Decision (MDPs) when the learner has access to a generative model. The advantages of state-of-the-art algorithms are discussed and illustrated.
  arXiv  Detail & Related papers  (2020-09-28T15:22:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.