7 papers
EvoScene-VLA: Evolving Scene Beliefs Inside the Action Decoder for Chunked Robot Control
Chushan Zhang, Ruihan Lu, Jinguang Tong +3
Chunked vision-language-action (VLA) policies generate multi-step actions from one observation and typically re-observe after executing the action chunk. Earlier actions change the…
Wasserstein Distributionally Robust Regret Optimization for Reinforcement Learning from Human Feedback
Yikai Wang, Shang Liu, Jose Blanchet
Reinforcement learning from human feedback (RLHF) is a central post-training tool for aligning large language models, but its training reward is only a learned proxy for true human…
Calibrating conditional risk
Andrey Vasilyev, Yikai Wang, Xiaocheng Li +1
We introduce and study the problem of calibrating conditional risk, which involves estimating the expected loss of a prediction model conditional on input features. We analyze this…
3D-IDE: 3D Implicit Depth Emergent
Chushan Zhang, Ruihan Lu, Jinguang Tong +2
Leveraging 3D information within Multimodal Large Language Models (MLLMs) has recently shown significant advantages for indoor scene understanding. However, existing methods, inclu…
Risk Profiling and Modulation for LLMs
Yikai Wang, Xiaocheng Li, Guanting Chen
Large language models (LLMs) are increasingly used for decision-making tasks under uncertainty; however, their risk profiles and how they are influenced by prompting and alignment…
Approximate MAP Estimation for Pairwise Potentials via Baker's Technique
Yi-Kai Wang
The theoretical models providing mathematical abstractions for several significant optimization problems in machine learning, combinatorial optimization, computer vision and statis…
Approximate Capacities of Two-Dimensional Codes by Spatial Mixing
Yi-Kai Wang, Yitong Yin, Sheng Zhong
We apply several state-of-the-art techniques developed in recent advances of counting algorithms and statistical physics to study the spatial mixing property of the two-dimensional…