1

Benchmarking Multi-national Value Alignment for Large Language Models
Do Large Language Models (LLMs) hold positions that conflict with your country’s values? Occasionally they do! However, existing works …
Cite
PKU-safeRLHF: Towards Multi-level Safety Alignment for LLMs with Human Preference
In this study, we introduce the safety human preference dataset, PKU-SafeRLHF, designed to promote research on safety alignment in …
PDF Cite Code DOI
Differentiable Information Enhanced Model-Based Reinforcement Learning
Differentiable environments have heralded new possibilities for learning control policies by offering rich differentiable information …
Cite
Distributed Policy Space Response Oracles in Two-Player Zero-Sum Games
Policy space response oracle (PSRO) is a population-based algorithm that can be used to solve two-player zero-sum games. In the PSRO …
PDF Cite
Magnetic Preference Optimization: Achieving Last-iterate Convergence for Language Model Alignment
Self-play methods have demonstrated remarkable success in enhancing model capabilities across various domains. In the context of …
PDF Cite
Rat: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviors
Evaluating deep reinforcement learning (DRL) agents against targeted behavior attacks is critical for assessing their robustness. These …
Cite
Towards Efficient Collaboration Via Graph Modeling In Reinforcement Learning
In multi-agent reinforcement learning, a commonly considered paradigm is centralized training with decentralized execution. However, in …
Cite
Differentiable Information Enhanced Model-Based Reinforcement Learning
Differentiable environments have heralded new possibilities for learning control policies by offering rich differentiable information …
PDF Cite
Falcon: Fast Visuomotor Policies via Partial Denoising
Diffusion policies are widely adopted in complex visuomotor tasks for their ability to capture multimodal action distributions. …
PDF Cite
Sequence to Sequence Reward Modeling: Improving RLHF by Language Feedback
Aligning the behavior of Large language models (LLMs) with human intentions and values remains a critical challenge. Reinforcement …
PDF Cite