相关论文: Quantifying and Optimizing Global Faithfulness in …

Counterfactual learning to rank (CLTR) can be risky and, in various circumstances, can produce sub-optimal models that hurt performance when deployed. Safe CLTR was introduced to mitigate these risks when using inverse propensity scoring to…

Inferring reward functions from demonstrations and pairwise preferences are auspicious approaches for aligning Reinforcement Learning (RL) agents with human intentions. However, state-of-the art methods typically focus on learning a single…

Humans are capable of attributing latent mental contents such as beliefs or intentions to others. The social skill is critical in daily life for reasoning about the potential consequences of others' behaviors so as to plan ahead. It is…

Human-AI complementarity is the claim that a human supported by an AI system can outperform either alone in a decision-making process. Since its introduction in the humanAI interaction literature, it has gained traction by generalizing the…

We study cautious reasoning in finite sequential games played by agents with perfect recall. Our contribution lies in formulating a definition of prudent rationalizability (Heifetz et al. 2021, BEJTE) as an iterative reduction procedure of…

We introduce the Principled Reasoning and Acting (PRAct) framework, a novel method for learning and enforcing action principles from trajectory data. Central to our approach is the use of text gradients from a reflection and optimization…

Spatial public goods games model collective dilemmas where individual payoffs depend on population-level strategy configurations. Most existing studies rely on evolutionary update rules or value-based reinforcement learning methods. These…

Persona-assigned large language models (LLMs) are used in domains such as education, healthcare, and sociodemographic simulation. Yet, they are typically evaluated only in short, single-round settings that do not reflect real-world usage.…

We consider the problem of approximate belief-state monitoring using particle filtering for the purposes of implementing a policy for a partially-observable Markov decision process (POMDP). While particle filtering has become a widely-used…

As machine learning becomes more widespread and is used in more critical applications, it's important to provide explanations for these models, to prevent unintended behavior. Unfortunately, many current interpretability methods struggle…

There have been many recent advances on provably efficient Reinforcement Learning (RL) in problems with rich observation spaces. However, all these works share a strong realizability assumption about the optimal value function of the true…

We study how well large language models (LLMs) explain their generations through rationales -- a set of tokens extracted from the input text that reflect the decision-making process of LLMs. Specifically, we systematically study rationales…

Justified representation (JR) and extended justified representation (EJR) are well-established proportionality axioms in approval-based multiwinner voting. Both axioms are always satisfiable, but they rely on a fixed quota (typically Hare…

We present a general framework for evolutionary learning to emergent unbiased state representation without any supervision. Evolutionary frameworks such as self-play converge to bad local optima in case of multi-agent reinforcement learning…

Conditional language models still generate unfaithful output that is not supported by their input. These unfaithful generations jeopardize trust in real-world applications such as summarization or human-machine interaction, motivating a…

Standard reinforcement learning from human feedback (RLHF) trains a reward model on pairwise preference data and then uses it for policy optimization. However, while reward models are optimized to capture relative preferences, existing…

This paper examines the effect of real-time, personalized alignment of a robot's reward function to the human's values on trust and team performance. We present and compare three distinct robot interaction strategies: a non-learner strategy…

Reinforcement learning algorithms rely on carefully engineering environment rewards that are extrinsic to the agent. However, annotating each environment with hand-designed, dense rewards is not scalable, motivating the need for developing…

Reward modeling has become a cornerstone of aligning large language models (LLMs) with human preferences. Yet, when extended to subjective and open-ended domains such as role play, existing reward models exhibit severe degradation,…

A core challenge for faithful LLM role-playing is sustaining consistent characterization throughout long, open-ended dialogues, as models frequently fail to recall and accurately apply their designated persona knowledge without explicit…