A new speculative inference system speeds up diffusion VLAs to 19.1 ms average latency (3.04x faster) on LIBERO by replacing most full 58 ms inferences with 7.8 ms draft rounds while preserving task performance.
hub Baseline reference
Diffusion policy: Visuomotor policy learning via action diffusion
Baseline reference. 50% of citing Pith papers use this work as a benchmark or comparison.
15
Pith papers citing it
Baseline
50% of classified citations
hub tools
citation-role summary
baseline 3
background 2
method 1
citation-polarity summary
representative citing papers
Proposes Rodrigues Network using a learnable Neural Rodrigues Operator to add kinematic inductive biases for improved robot action learning and prediction.
Averaging and temporally interpolating text latents in VLAs enables 83% success on novel task combinations in the libero-ood benchmark where SOTA models achieve under 15%.
Adaptive Q-Chunking selects optimal action chunk sizes at each state via normalized advantage comparisons to outperform fixed chunk sizes in offline-to-online RL on robot benchmarks.
DIAL decouples intent from action in end-to-end VLAs using a latent visual foresight bottleneck and two-stage training, reaching SOTA on RoboCasa with 10x fewer demonstrations and zero-shot real-world transfer.
Q-chunking improves offline-to-online RL sample efficiency on long-horizon sparse-reward manipulation tasks by applying action chunking to TD learning.
RoboTwin 2.0 automates diverse synthetic data creation for dual-arm robots via MLLMs and five-axis domain randomization, leading to 228-367% gains in manipulation success.
GAF creates 4D dynamic scene models by adding motion to 3D Gaussians, enabling better reconstruction and 7.3% higher success in robotic tasks.
BAC accelerates transformer-based Diffusion Policy up to 3x by block-level adaptive feature caching using an Adaptive Caching Scheduler and Bubbling Union Algorithm to control error propagation.
Real-time chunking (RTC) allows diffusion- and flow-based action chunking policies to execute smoothly and asynchronously, maintaining high success rates on dynamic tasks even with significant inference latency.
DreamPolicy integrates an autoregressive diffusion world model with policy learning to produce a single scalable policy that generalizes to unseen composite terrains for humanoid locomotion.
HybridVLA unifies diffusion and autoregression in a single VLA model via collaborative training and ensemble to raise robot manipulation success rates by 14% in simulation and 19% in real-world tasks.
UVA learns a joint video-action latent representation with decoupled diffusion decoding heads, enabling a single model to perform accurate fast policy learning, forward/inverse dynamics, and video generation without performance loss versus task-specific methods.
OpenVLA-OFT fine-tuning boosts LIBERO success rate from 76.5% to 97.1%, speeds action generation 26x, and outperforms baselines on real bimanual dexterous tasks.
Systematic tests of VLM backbones, policy architectures, and cross-embodiment data yield RoboVLMs that set new SOTA on robot manipulation benchmarks while requiring few manual designs.
citing papers explorer
-
Realtime-VLA FLASH: Speculative Inference Framework for Diffusion-based VLAs
A new speculative inference system speeds up diffusion VLAs to 19.1 ms average latency (3.04x faster) on LIBERO by replacing most full 58 ms inferences with 7.8 ms draft rounds while preserving task performance.
-
Rodrigues Network for Learning Robot Actions
Proposes Rodrigues Network using a learnable Neural Rodrigues Operator to add kinematic inductive biases for improved robot action learning and prediction.
-
VLAs are Confined yet Capable of Generalizing to Novel Instructions
Averaging and temporally interpolating text latents in VLAs enables 83% success on novel task combinations in the libero-ood benchmark where SOTA models achieve under 15%.
-
Adaptive Q-Chunking for Offline-to-Online Reinforcement Learning
Adaptive Q-Chunking selects optimal action chunk sizes at each state via normalized advantage comparisons to outperform fixed chunk sizes in offline-to-online RL on robot benchmarks.
-
DIAL: Decoupling Intent and Action via Latent World Modeling for End-to-End VLA
DIAL decouples intent from action in end-to-end VLAs using a latent visual foresight bottleneck and two-stage training, reaching SOTA on RoboCasa with 10x fewer demonstrations and zero-shot real-world transfer.
-
Reinforcement Learning with Action Chunking
Q-chunking improves offline-to-online RL sample efficiency on long-horizon sparse-reward manipulation tasks by applying action chunking to TD learning.
-
RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation
RoboTwin 2.0 automates diverse synthetic data creation for dual-arm robots via MLLMs and five-axis domain randomization, leading to 228-367% gains in manipulation success.
-
GAF: Gaussian Action Field as a 4D Representation for Dynamic World Modeling in Robotic Manipulation
GAF creates 4D dynamic scene models by adding motion to 3D Gaussians, enabling better reconstruction and 7.3% higher success in robotic tasks.
-
Block-wise Adaptive Caching for Accelerating Diffusion Policy
BAC accelerates transformer-based Diffusion Policy up to 3x by block-level adaptive feature caching using an Adaptive Caching Scheduler and Bubbling Union Algorithm to control error propagation.
-
Real-Time Execution of Action Chunking Flow Policies
Real-time chunking (RTC) allows diffusion- and flow-based action chunking policies to execute smoothly and asynchronously, maintaining high success rates on dynamic tasks even with significant inference latency.
-
DreamPolicy: A Unified World-model Policy for Scalable Humanoid Locomotion
DreamPolicy integrates an autoregressive diffusion world model with policy learning to produce a single scalable policy that generalizes to unseen composite terrains for humanoid locomotion.
-
HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Model
HybridVLA unifies diffusion and autoregression in a single VLA model via collaborative training and ensemble to raise robot manipulation success rates by 14% in simulation and 19% in real-world tasks.
-
Unified Video Action Model
UVA learns a joint video-action latent representation with decoupled diffusion decoding heads, enabling a single model to perform accurate fast policy learning, forward/inverse dynamics, and video generation without performance loss versus task-specific methods.
-
Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
OpenVLA-OFT fine-tuning boosts LIBERO success rate from 76.5% to 97.1%, speeds action generation 26x, and outperforms baselines on real bimanual dexterous tasks.
-
What Matters in Building Vision-Language-Action Models for Generalist Robots
Systematic tests of VLM backbones, policy architectures, and cross-embodiment data yield RoboVLMs that set new SOTA on robot manipulation benchmarks while requiring few manual designs.