4arXiv cs.AI (Artificial Intelligence)·
UR-VC: Unsupervised correction of time-derived progress labels for robot learning
UR-VC is a training-free offline method for correcting noisy time-derived progress labels used in robot learning, which fail to capture regressions caused by slips or failed grasps in contact-rich manipulation. The approach retrieves similar states across demonstration episodes and aggregates their timestamps to produce corrected progress estimates, requiring no manual annotations or additional value models. The method is evaluated on real bimanual cloth folding tasks and used to construct advantage labels for vision-language-action (VLA) policy training, showing positive trends in task success.
Related guides (1)
Related events (8)
RECALL: Active continual learning for Vision-Language-Action models via uncertainty-guided recovery data collection
Researchers propose RECALL, an active continual learning paradigm for Vision-Language-Action (VLA) robot models that uses uncertainty-guided data collection to target states where the policy struggles, rather than passively collecting demonstrations after failures. The paper demonstrates improved fine-tuning efficiency over passive imitation learning but identifies catastrophic forgetting as a key challenge when incorporating recovery data. The authors evaluate continual learning mitigations including replay-based data mixing and elastic weight consolidation, characterizing tradeoffs between plasticity and retention in large autoregressive robot policies.
FORCE: Efficient RL fine-tuning for Vision-Language-Action models via value-calibrated warm-up and self-distillation
Researchers introduce FORCE, a 3-stage reinforcement learning fine-tuning framework for Vision-Language-Action (VLA) models that addresses sample inefficiency caused by unstable Q-functions and low-quality exploration data. The framework uses a Value-Calibrated Warm-Up phase followed by Q-function-filtered policy updates, eliminating the need for costly human interventions during training. Evaluated on simulation and real-world robotic tasks, FORCE achieves a 79% absolute improvement in task success rates, outperforms prior RL methods by 10%, and accelerates training by 32.5%.
HABC: Hierarchical Advantage Weighting for Online RL Fine-Tuning of Vision-Language-Action Policies
Researchers introduce Hierarchical Advantage-Weighted Behavior Cloning (HABC), a method for fine-tuning pretrained Vision-Language-Action (VLA) policies via online RL using only sparse binary episode outcomes. HABC trains separate critic heads for viability and efficiency objectives, combines them via a state-adaptive gate, and applies intervention-aware credit assignment to avoid incorrect supervision across human-intervention boundaries. On three contact-rich bimanual real-robot tasks, HABC improves success rates from SFT baselines of 36%, 44%, and 12% to 92%, 88%, and 38% respectively. The work addresses a fundamental credit assignment problem in robot learning from sparse outcome signals.
VRRL: Reinforcement learning framework for visually grounded self-reflection in vision-language models
Researchers propose VRRL, a reinforcement learning training framework designed to improve self-reflection in large vision-language models (LVLMs) by ensuring corrections are grounded in visual inputs rather than purely textual reasoning. The framework uses two mechanisms: random masking of trajectory prefixes to emphasize error recovery, and buffered roll-ins from an experience replay buffer to expose models to diverse failure states. Evaluated on visual grounding tasks (tables, charts) and spatial navigation benchmarks, VRRL substantially improves out-of-distribution accuracy over standard RL and reflection-oriented fine-tuning baselines. The work addresses a concrete failure mode where LVLMs lose visual grounding during chain-of-thought reflection.
TREAD: VLM-based re-labelling framework improves robot policy generalization via dataset augmentation
TREAD (Task Robustness via Re-Labelling Vision-Action Robot Data) is a scalable framework that uses pretrained Vision-Language Models to augment existing robotics datasets without new data collection. The approach decomposes demonstrations into sub-tasks, segments videos accordingly, and generates linguistically diverse instruction labels, enriching language-action pair diversity. Evaluations on the LIBERO benchmark show improved generalization to novel tasks and goals, addressing a key limitation of current robot learning policies.
Sony and University Researchers Train Robots To Learn Without Catastrophic Forgetting
Researchers from UT Austin, UCLA, Nanyang Technological University, and Sony developed a sequential fine-tuning recipe combining LoRA and on-policy reinforcement learning (GRPO) to reduce catastrophic forgetting in vision-language-action (VLA) models for robotics. Applied to the OpenVLA-OFT model on the LIBERO benchmark, the method achieved 81.2% success on libero-spatial tasks with near-zero forgetting (0.3 percentage point drop), outperforming established continual learning baselines including Dark Experience Replay and Elastic Weight Consolidation. The approach requires no replay of prior task data and also showed modest generalization to unseen tasks. The authors note the method has not yet been tested outside robotics simulation contexts.
VLK: Synthetic vision-language-kinematics supervision enables humanoid loco-manipulation from egocentric observations
Researchers introduce a pipeline that generates 48,000 paired vision-language-kinematics trajectories synthetically using 3D Gaussian Splatting to reconstruct indoor scenes, bypassing the need for expensive human-annotated robot data. A VLK policy trained on this data predicts whole-body kinematic trajectories from egocentric images and language instructions, which a whole-body tracker converts to physical actions. The approach is validated on a Unitree G1 humanoid performing navigation and object transport, demonstrating viable sim-to-real transfer for perception-based loco-manipulation.
VERITAS: Visual verification enables inference-time steering and autonomous improvement for robot policies
Researchers introduce VERITAS, a generator-verifier framework pairing a pre-trained generalist robot policy with a gradient-free visual verifier to steer actions at inference time without additional training. Verified rollouts are also used for offline self-improvement via fine-tuning, achieving performance gains comparable to expert demonstrations but without human intervention. The work demonstrates that inference-time verification is a scalable mechanism for autonomous policy improvement during deployment.