FORGE-Tree: Diffusion-Forcing Tree Search for Long-Horizon Robot Manipulation
Yanjia Huang, Shuo Liu, Sheng Liu, Qingxiao Xu, Mingyang Wu, Xiangbo Gao, Zhengzhong Tu

TL;DR
FORGE-Tree introduces a novel diffusion-based tree search method for long-horizon robot manipulation, improving success rates by local trajectory refinement and selective search, addressing challenges of error accumulation and compute allocation.
Contribution
It presents FORGE-Tree, a new plug-in control layer combining diffusion forcing and Monte Carlo tree diffusion for enhanced long-horizon manipulation.
Findings
Improves success rate by up to 17.2 percentage points over baselines.
Scales performance with search budget while preserving executed prefix.
Maintains gains under limited compute budgets, especially on long-horizon tasks.
Abstract
Long-horizon robot manipulation tasks remain challenging for Vision-Language-Action (VLA) policies due to drift and exposure bias, often denoise the entire trajectory with fixed hyperparameters, causing small geometric errors to compound across stages and offering no mechanism to allocate extra test-time compute where clearances are tight. To address these challenges, we introduce FORGE-Tree, a plug-in control layer that couples a stage-aligned Diffusion Forcing (DF) head with test-time Monte Carlo Tree Diffusion (MCTD). With a frozen VLA encoder, DF aligns timesteps to subtask stages; during inference we partially denoise only a target segment while keeping other tokens frozen, turning trajectory refinement into a sequence of local edits. We then apply Monte Carlo Tree Diffusion to select the next segment to refine. A scene graph supplies priors for expansion and geometry…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
