LTLDiff: Finite Linear Temporal Logic-Guided Data Generation and Diffusion Policies for Multi-agent Robotic Manipulation
针对多机器人协作操作中的同步、动作顺序和协调问题，提出LTLDiff框架，通过有限线性时序逻辑指导数据生成和扩散策略学习，提高任务成功率。
0 citationsRead paper
Selected work
Representative Papers
Ripple-Pivot Search: Active Parallel Decoding for Diffusion Large Language Models
This work addresses the inefficiency of existing parallel decoding strategies in diffusion language models, which overlook the potential of early deterministic decisions to enhance global decoding efficiency. The authors propose a training-free active parallel decoding method that, for the first time, identifies and leverages a “ripple effect” during decoding: by detecting medium-entropy “pivot” positions, prospectively evaluating their impact on downstream uncertainty, and dynamically scheduling optimal decoding paths using KV cache management. Evaluated across three diffusion language models and four benchmarks spanning reasoning and code generation, the approach achieves 4–10× end-to-end speedup (up to 18× in peak cases) while preserving generation quality and consistently outperforming prior state-of-the-art baselines by up to 5.49% in accuracy across most settings.
0 citationsRead paper
Recent publications
Latest Papers
LTLDiff: Finite Linear Temporal Logic-Guided Data Generation and Diffusion Policies for Multi-agent Robotic Manipulation
针对多机器人协作操作中的同步、动作顺序和协调问题，提出LTLDiff框架，通过有限线性时序逻辑指导数据生成和扩散策略学习，提高任务成功率。
0 citationsRead paper
Ripple-Pivot Search: Active Parallel Decoding for Diffusion Large Language Models
This work addresses the inefficiency of existing parallel decoding strategies in diffusion language models, which overlook the potential of early deterministic decisions to enhance global decoding efficiency. The authors propose a training-free active parallel decoding method that, for the first time, identifies and leverages a “ripple effect” during decoding: by detecting medium-entropy “pivot” positions, prospectively evaluating their impact on downstream uncertainty, and dynamically scheduling optimal decoding paths using KV cache management. Evaluated across three diffusion language models and four benchmarks spanning reasoning and code generation, the approach achieves 4–10× end-to-end speedup (up to 18× in peak cases) while preserving generation quality and consistently outperforming prior state-of-the-art baselines by up to 5.49% in accuracy across most settings.
0 citationsRead paper