The Unexpired Plan: A Free Monitor for Accelerated Diffusion Policies
cs.RO
Submitted: 2026-10-05
Updated: 2026-10-05
Sources
- Exploring the Limitations of Behavior Cloning for Autonomous Driving
- DeepCache: Accelerating Diffusion Models for Free
- WorldCache: Content-Aware Caching for Accelerated Video World Models
- MagCache: Fast Video Generation with Magnitude-Aware Cache
- Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model
- Neural Introspection Gating for Adaptive KV-Cache Reuse in Vision-Language-Action Models
- ActionCache: Training-Free Acceleration for Vision-Language-Action Models with Action Caching and Refinement
- C$^3$ache: Accelerating World Action Models with Cross Inference Chunk Cache
- Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Models
- DeeR-VLA: Dynamic Inference of Multimodal Large Language Models for Efficient Robot Execution
- Consistency Policy: Accelerated Visuomotor Policies via Consistency Distillation
- Learning-to-Cache: Accelerating Diffusion Transformer via Layer Caching
- The Gate, Not the Cache: Gate Provenance Bounds the Closed-Loop Reliability of Training-Free VLA Token Skipping
- Just Noticeable Difference Modeling for Token Compression in Vision-Language-Action Models
- Denoising Tells When to Replan: Denoising-Variance Adaptive Chunking for Flow-Based Robot Policies
- Adaptive Caching for Faster Video Generation with Diffusion Transformers
- Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware
- Bidirectional Decoding: Improving Action Chunking via Guided Test-Time Sampling
- Unpacking Failure Modes of Generative Policies: Runtime Monitoring of Consistency and Progress
- Fast Inference from Transformers via Speculative Decoding
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies