Plan First, Diffuse Later: Extrinsic Graph Guidance for Long-Horizon Diffusion Planning
For robotic planning and sequential decision-making, XDiffuser offers a more efficient way to achieve global coherence in long-horizon tasks by offloading exploration to classical algorithms.
XDiffuser improves long-horizon planning by using extrinsic search over a state-space graph to guide diffusion denoising, outperforming diffusion baselines on long-horizon tasks with large gains in low-quality data regimes and unseen tasks like multi-agent coordination and TSP-style reasoning.
Compositional diffusion models offer a promising route to long-horizon planning by denoising multiple overlapping sub-trajectories while ensuring that together they constitute a global solution. However, enforcing local behavior over long chains is often insufficient for a coherent global structure to emerge. Recent works tackle this limitation through intrinsic search, which explores multiple paths during the denoising process. While intrinsic search improves global coherence, it comes at the cost of repeated evaluations of an already compute-heavy model. In this work, we argue that extrinsic search, performed outside the denoising process, offers a more effective mode of exploration for long-horizon planning while naturally enabling the use of classical algorithms to solve unseen combinatorial tasks at test time. Our eXtrinsic search-guided Diffuser (XDiffuser) first computes a plan over a state-space graph -- serving as a lightweight local connectivity oracle for the diffusion model. The plan is then used to guide denoising for a single trajectory, effectively offloading the burden of exploration. XDiffuser outperforms diffusion-based baselines on long-horizon tasks, with particularly large gains in the low-quality data regime and on unseen tasks beyond goal-reaching, including multi-agent coordination and TSP-style reasoning. Project website: https://yanivhass.github.io/XDiffuser-site/