Wednesday, November 11, 2026
24 papers added this day. New papers stay for 14 days.
Dexterous manipulation 1
Imitation learning and data collection 3
Robot foundation models 9
RL and sim-to-real 3
Humanoids and mobile manipulation 5
Planning and control 3
One Policy for 100 Household Tasks: Our Solution for the 2026 BEHAVIOR Challenge
November 2026
One pi0.5-based VLA checkpoint, trained for 200k steps on 20,000 demonstrations, covers all 100 BEHAVIOR household tasks in simulation and scores 0.392 Q; on the 50 tasks shared with 2025, Q rises from 0.260 to 0.371. Routing 14 hard tasks to a GPT-6 Astra coding agent that practiced beforehand lifts the public submission to 0.446 Q.
From Prediction to Action: Learning Physical Context for Deformable Linear Object Manipulation
November 2026
For shaping cables and ropes, knowing the stiffness and friction helps predict the next shape but not choose the action; a latent of recent interaction history does. Trained only in simulation, the history-conditioned policy brings real ropes within 10 mm of the target shape in 88.3% of rollouts on a UR5, against 48.3% for a diffusion-policy baseline.
RedirectMimic: Local Motion Content Redaction for Humanoid Whole-Body Control
November 2026
Fine-tune a humanoid motion tracker so it swaps out selected arm movements while still being fed the original reference, by computing tracking rewards against a rule-edited reference. In Unitree G1 simulation, overhead-arm content drops to 2.7% of rollouts with 94.2% joint success.
LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
November 2026
A world action model that adds a latent expert between video prediction and action generation, trained to focus on robot-object interaction, plus a progress value that decides how many predicted actions to execute. Pretrained on over 14k hours, LaST0.5 reaches 99.5% on LIBERO and 63.8% on four real bimanual tasks against 52.5% for π0.5.
ReACT: Reflection-Consistent Modulation and Cross-Load Consistency for Humanoid Velocity Tracking Under Asymmetric Loads
November 2026
A humanoid carrying an object in one hand drifts sideways and in yaw. ReACT feeds an estimated load wrench into a mirror-symmetric policy and trains loaded and unloaded rollouts to move alike, and a Unitree G1 carrying 1.5 kg in one hand then reaches the goal collision-free in 40/40 navigation trials where the FALCON baseline fails all 40.
CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation
November 2026
CIALLO is an action codec pretrained on eight embodiments that maps actions, visual changes and language into one shared intent space, and adapts to a new robot from a single recording of at most 10 seconds. A Qwen-based policy on top reaches 76.7% success on real SO-101 and Piper arms with 20 demonstrations per task, against 70.0% for fully finetuned pi0.
Few-Shot Residual Adaptation of Pretrained Latent World Models
November 2026
Small residual adapters, trained self-supervised on three interaction episodes, correct a frozen latent world model both inside its predictor and at its output. On a real Franka pushing task, success recovers from 0% to 90% under a center-of-mass shift and from 3.3% to 100% under changed friction.
ExRL: Sample-Efficient Online Adaptation of Long-Chunk Policies via Execution Length Control
November 2026
Instead of using RL to change the actions of a long-chunk policy, ExRL keeps the policy frozen and learns only how many actions of each predicted chunk to execute before replanning. On a real arm it lifts a frozen pi0.5 from at most 9/20 to 17/20 on cube stacking, with about ten minutes of interaction.
Robo-MLT: Dual-System Robotic Manipulation for Long-Horizon Medical Laboratory Tasks
November 2026
A dual-arm robot runs blood gas analysis and tube sorting by keeping the protocol as an explicit skill list, with a small multiframe monitor that advances or replans it while a VLA executes each skill. Task success is 82% and 70%, 30 and 22 points above pi0.5.
World In-Context Learning for Agentic Robot Manipulation
November 2026
A frozen foundation model learns each manipulation task from one demonstration by rehearsing in a simulator and writing what it learns into a text Task Memory instead of weights. RoboScript reaches 83.4% mean success on the 50 RoboTwin 2.0 tasks, against 61.7% for the best trained policy, and the memories transfer to a real dual-arm robot.
SWAM: Scaling World Action Models with Cross-Modal Alignment
November 2026
A study of how to pretrain world action models on about 47,000 hours of general video, egocentric video, UMI and robot demonstrations while keeping the video model's language-video alignment. The resulting recipe reaches 61.56% on RoboTwin under randomized scenes and 40.00% on three real bimanual tasks, against 23.33% for the strongest baseline.
RAIL: Representation-Aligned Imitation Learning for Student-Compatible Teacher Policies
November 2026
Train the privileged RL teacher on a latent space that the student's observations can also reach, so the student runs the same policy without distillation. In four simulated tasks the teacher-student gap nearly vanishes (success 1.00, 1.00 and 0.98) and the student beats RL from raw observations on the two pixel-based tasks.
Flip Anything with Agentic Robotics Self-Learning
November 2026
A coding agent writes and refines a graph-based bimanual flipping policy purely by trial in simulation, with no demonstrations. Over self-learning iterations success on 24 objects rises from 24.7% to 95.0%, and the code runs zero-shot on real YAM arms.
Distilling Instructions into Weights: Closed-Form Edits Fix a Robot Policy's Misreading
November 2026
A zero-shot pi0.5 policy misreads "put the cube then the banana in the bowl" and drops the cube early. Copying the internal effect of a clearer rewording into the VLM's MLP weights in closed form, with no gradients, raises held-out simulation success from 27/128 to 106/128.
Design of a hybrid tendon-linkage finger with coupled and adaptive motion
November 2026
A robot finger with four joints driven by three motors, where the fingertip joint follows the middle joint through a spring-loaded four-bar linkage and bends further once contact blocks the middle joint. The built finger tracks the predicted coupling with 2.54 degrees mean error and presses with 4.56 N at the fingertip.
PAIR: Proprioception-Aligned Image-Ray Representations for Viewpoint-Robust Manipulation
November 2026
PAIR feeds camera rays into the vision encoder of an imitation policy and adds a contrastive loss that pulls together observations with similar robot proprioception, so features stay consistent when the camera moves. On a real Franka with cameras repositioned every rollout, final task success rises 26.7 points over plain behavior cloning.
Accelerated Reinforcement Learning for Soft Robotics: Hours Instead of Years
November 2026
A 6,924-parameter reduced-order model of a pneumatic soft arm, fit in 5 minutes from about two hours of hardware data, serves as the RL training environment. The policy trains in 3 hours, deploys without fine-tuning and tracks targets with 40% lower error than policies trained on RNN or transformer models.
Data-Driven Precision Control of Soft Helicoid Robots
November 2026
A 5D reduced-order model of a tendon-driven soft trunk robot is learned from two hours of motion-capture data and used in 50 Hz model predictive control. On hardware it keeps the maximum tracking error to 5 cm across 2D and 3D tracks, against 10-38 cm for linear, Koopman and RL baselines, and also avoids virtual obstacles.
Streaming Failure Forecasting: VLA Failure Monitors That Keep Learning After Deployment
November 2026
A failure monitor for VLA policies that refits only its last linear layer after every episode from the success/failure outcome. It ranks failures better than the same monitor kept frozen on 4 of 5 simulated drift suites and on 434 recorded real SO-101 runs of four VLAs, at 2.18 ms per policy call on a Jetson CPU core.
HINT-VLM: Humanoid Interactive Navigation over Rough Terrain with Embodiment-Grounded Vision-Language Model
November 2026
A Unitree G1 humanoid decides before contact whether to push an obstacle aside or walk around it, using a VLM's movability guess corrected by a Gaussian process fit to simulated pushes. It reaches the goal in 95.5-100% of simulated rough-terrain trials and cuts real-world navigation distance from 16.8 m to 10.3 m.
From Festival Video to the Stage: A Sim-to-Real Ladder for Long-Horizon Expressive Whole-Body Dance on a Commercial Humanoid
November 2026
A video-to-robot pipeline puts a 43-second Awa Odori festival dance on a Unitree G1, with each of eleven policy versions graded on hardware. Drift-aware observation noise and scripted edits of the reference motion take it from early falls to a continuous 9.7-minute rope-free performance, all trained on one RTX 5070.
Control-Aware Command Generation for Asynchronous Control and Action Inference with Vision-Based Policies towards Fast Execution
November 2026
A tracking-control layer turns the jumpy action chunks of an asynchronously running policy into smooth, acceleration-limited joint targets, and skips actions when successive predictions agree. On a bimanual ALOHA with ACT, tasks take 53-63% of the synchronous execution time and pen-storage success rises from 30% to 83%.
Zero-Shot VLA Adaptation with Retrieved Guidance
November 2026
RAGu retrieves demonstrations from mismatched datasets such as DROID, warps them into the target scene and uses them to steer a frozen VLA's flow-matching denoising, with no training. Real-world success of pi0.5 rises from 20% to 80%.
DreamVLA: From Generated Human Motions to Humanoid Vision-Language-Action Policies
November 2026
Generated human motions guide an RL expert in simulation, whose successful rollouts and recoveries from the student's failures train a GR00T policy for a Unitree G1, with no task-specific teleoperation. Three DAgger rounds lift simulated picking from 0% to 54%, while the same amount of plain expert data stays at 0%.
Summaries are written automatically from each paper's text. Figures and videos come from the paper and its project page; all credit goes to the authors.