相关论文: Learning Correlated Reward Models: Statistical Bar…

A common approach to personalization in large language models (LLMs) is to incorporate a subset of the user memory into the prompt at inference time to guide the model's generation. Existing methods select these subsets primarily using…

Reward models (RMs) are at the crux of successfully using RLHF to align pretrained models to human preferences, yet there has been relatively little study that focuses on evaluation of those models. Evaluating reward models presents an…

Rewards serve as proxies for human preferences and play a crucial role in Reinforcement Learning from Human Feedback (RLHF). However, if these rewards are inherently imperfect, exhibiting various biases, they can adversely affect the…

Preference alignment in Large Language Models (LLMs) has significantly improved their ability to adhere to human instructions and intentions. However, existing direct alignment algorithms primarily focus on relative preferences and often…

Reinforcement learning based fine-tuning of large language models (LLMs) on human preferences has been shown to enhance both their capabilities and safety behavior. However, in cases related to safety, without precise instructions to human…

We study a dynamic generalization of stochastic rationality in consumer behavior, the Dynamic Random Utility Model (DRUM). Under DRUM, a consumer draws a utility function from a stochastic utility process and maximizes this utility subject…

Human cooperation depends on how accurately we infer others' motives--how much they value fairness, generosity, or self-interest from the choices they make. We model that process in binary dictator games, which isolate moral trade-offs…

Machine learning systems embed preferences either in training losses or through post-processing of calibrated predictions. Applying information design methods from Strack and Yang (2024), this paper provides decision problem agnostic…

Aligning large language models (LLMs) with human preferences is a critical challenge in AI research. While methods like Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) are widely used, they often…

Probabilistic user modeling is essential for building machine learning systems in the ubiquitous cases with humans in the loop. However, modern advanced user models, often designed as cognitive behavior simulators, are incompatible with…

Reinforcement learning (RL) faces challenges in evaluating policy trajectories within intricate game tasks due to the difficulty in designing comprehensive and precise reward functions. This inherent difficulty curtails the broader…

This paper examines the Random Utility Model (RUM) in repeated stochastic choice settings where decision-makers lack full information about payoffs. We propose a gradient-based learning algorithm that embeds RUM into an online…

Recent progress in strengthening the capabilities of large language models has stemmed from applying reinforcement learning to domains with automatically verifiable outcomes. A key question is whether we can similarly use RL to optimize for…

Reinforcement Learning from Human Feedback (RLHF) has emerged as a popular paradigm for aligning models with human intent. Typically RLHF algorithms operate in two phases: first, use human preferences to learn a reward function and second,…

Learning from human preferences is a cornerstone of aligning machine learning models with subjective human judgments. Yet, collecting such preference data is often costly and time-consuming, motivating the need for more efficient learning…

Despite the success of reinforcement learning from human feedback (RLHF) in aligning language models, current reward modeling heavily relies on experimental feedback data collected from human annotators under controlled and costly…

Reinforcement Learning from Human Feedback (RLHF) has achieved considerable success in aligning large language models (LLMs) by modeling human preferences with a learnable reward model and employing a reinforcement learning algorithm to…

Recent advances in generative AI have been driven by alignment techniques such as reinforcement learning from human feedback (RLHF). RLHF and related techniques typically involve constructing a dataset of binary or ranked choice human…

Reward models are central to aligning language models with human preferences via reinforcement learning (RL). As RL is increasingly applied to settings such as verifiable rewards and multi-objective alignment, RMs are expected to encode…

Aligning language models to human expectations, e.g., being helpful and harmless, has become a pressing challenge for large language models. A typical alignment procedure consists of supervised fine-tuning and preference learning. Most…