Fast Monte Carlo Tree Diffusion: 100x Speedup via Parallel Sparse Planning

2025-06-11

AI Summary

Diffusion models suffer from low inference efficiency in long-horizon trajectory planning, while Monte Carlo Tree Diffusion (MCTD) incurs substantial computational overhead due to sequential search and iterative denoising. To address these challenges, this paper proposes Fast-MCTD—a novel framework featuring two key innovations: (1) Parallel MCTD, which enables efficient concurrent rollouts via delayed parameter updates and redundancy-aware node selection; and (2) Sparse MCTD, which coarsens the temporal dimension of trajectories to reduce modeling complexity and drastically shrink the search space. Together, these components achieve over 100× speedup without compromising planning quality. Experiments on multi-task planning benchmarks demonstrate that Fast-MCTD attains up to 100× acceleration over standard MCTD, outperforms search-free diffusion planners in inference speed, and matches or exceeds their planning performance.

Technology Category

• Planning, Routing, and Scheduling: Planning with Markov Models (MDPs, POMDPs)
• Multiagent Systems: Multiagent Planning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Diffusion models have recently emerged as a powerful approach for trajectory planning. However, their inherently non-sequential nature limits their effectiveness in long-horizon reasoning tasks at test time. The recently proposed Monte Carlo Tree Diffusion (MCTD) offers a promising solution by combining diffusion with tree-based search, achieving state-of-the-art performance on complex planning problems. Despite its strengths, our analysis shows that MCTD incurs substantial computational overhead due to the sequential nature of tree search and the cost of iterative denoising. To address this, we propose Fast-MCTD, a more efficient variant that preserves the strengths of MCTD while significantly improving its speed and scalability. Fast-MCTD integrates two techniques: Parallel MCTD, which enables parallel rollouts via delayed tree updates and redundancy-aware selection; and Sparse MCTD, which reduces rollout length through trajectory coarsening. Experiments show that Fast-MCTD achieves up to 100x speedup over standard MCTD while maintaining or improving planning performance. Remarkably, it even outperforms Diffuser in inference speed on some tasks, despite Diffuser requiring no search and yielding weaker solutions. These results position Fast-MCTD as a practical and scalable solution for diffusion-based inference-time reasoning.

Problem

• Improve speed of Monte Carlo Tree Diffusion planning
• Reduce computational overhead in diffusion-based planning
• Enhance scalability for long-horizon reasoning tasks

Innovation

• Parallel MCTD enables parallel rollouts efficiently
• Sparse MCTD reduces rollout length via coarsening
• Combines tree search with diffusion for speedup