Overview

Robot policy learning studies how robots can turn camera, depth, tactile, and proprioceptive inputs into actions for tasks such as manipulation, locomotion, and navigation. Common approaches include behavior cloning from demonstrations, reinforcement learning in simulation or on hardware, and diffusion, flow-matching, or vision-language-action policies that generate action sequences; world models and hierarchical planners can add foresight and long-horizon structure. Researchers evaluate these methods in simulation suites such as LIBERO, CALVIN, RoboTwin, and navigation benchmarks, as well as on physical robots, with growing attention to memory, contact, safety, and recovery from failure. A central tension is how to make policies generalize across tasks, robots, and changing conditions while remaining data-efficient, reactive, reliable, and fast enough to deploy.

Latest papers

405 papers42 papers in the last four weeks, down 58% on the four weeks before. 0.3% of all new papers.

Oct 13, 2026

Interaction Latents as Actions for Contact-Rich Manipulation

Contact-rich manipulation requires robots to coordinate not only where to move, but also how contact and force should evolve throughout an interaction. Yet most robot policies represent actions primarily as motion trajectories, using tactile signals as inputs or auxiliary prediction targets without explicitly encoding contact dynamics in the action representation. We propose ILA (Interaction Latents as Actions), which uses decodable interaction latents that jointly capture the evolution of motion, contact, and force application as policy action units, providing a shared representation for action generation, evaluation, and offline improvement. We learn these latents through visually conditioned joint action-tactile reconstruction with a time-conditioned implicit decoder. Using them as flow-matching generation targets supports post-training of pretrained vision-language-action models (VLAs) and world action models (WAMs), as well as training flow-matching policies from scratch. A shared decoder maps generated latents to executable actions. For offline improvement, we learn a tactile-conditioned expert reference model from interaction latents encoded from expert demonstrations. The resulting expert-consistency rewards guide weighted learning from successful rollouts, while observation-compatible expert references guide correction of low-reward predictions from failed rollouts. Across three policy families, ILA improves success rates averaged over four real-world tasks by 21.0-36.5 percentage points relative to tactile-conditioned direct-action baselines. Ablations highlight the benefit of encoding contact evolution in action targets beyond tactile conditioning, while the same latent space supports expert-guided offline improvement. Project website: https://mrsecant.github.io/ILA/.
Oct 13, 2026

OneDemo: Unified Policy Learning and Data Generation from a Single Demonstration

Learning robot policies from demonstrations is often limited by the cost of collecting diverse, high-quality expert data. While an appealing alternative is to expand a single successful demonstration into a diverse set of successful trajectories, existing augmentation methods typically rely on manually designed transformation rules or complex trajectory planning, requiring substantial task-specific engineering and human effort. We introduce OneDemo, a reinforcement learning framework that learns to edit a single successful demonstration for new task configurations. We formulate demonstration editing as a structured reinforcement learning problem, in which an editor learns, from sparse task-success rewards, both when to transition between object-centric trajectory segments and how to geometrically adapt each segment. The learned editor serves both as a task policy and as a reusable demonstration generator. Across 15 Meta-World tasks, OneDemo reaches a 92.7% success rate within 53.7k environment interactions on average. Across 9 DexMimicGen tasks, it generates successful demonstrations with a 98.4% average success rate, and BC-RNN policies trained on the generated data achieve 99.6% average success at matched dataset scale. These results show that learning to edit a single successful demonstration provides a structured and effective basis for both policy learning and scalable demonstration generation.
Oct 13, 2026

Training Generalist Navigation Policies via Action Self-Distillation

Language provides a natural interface for humans to specify and reason about goals, but training language-conditioned robotic policies and reasoning typically requires expensive manual annotations. We propose Action Self-Distillation (ASD), a method for learning language-conditioned reasoning for robotic navigation without human-annotated language or reasoning labels. ASD uses an open-source VLM, which contains broad world knowledge and reasoning capabilities, as both a teacher network and a student policy. During training, the self-teacher receives privileged trajectory feedback and infers the underlying navigation goal together with goal-oriented reasoning. The student is then trained to condition on the inferred navigation goal, predict future ego-frame trajectory, and provide a description of the scene semantics, plan, and reasoning, from onboard observations alone. Across 13 challenging indoor and outdoor real-world navigation tasks, our method improves success rate over the strongest baseline by 30%30\% in indoor settings, 30%30\% on multi-stage tasks, and achieves 95%95\% success rate on outdoor tasks, validating that privileged trajectory feedback can enable self-distillation for scalable generalist navigation policies without human annotations.
Oct 12, 2026

MIRROR: Reflecting Hidden Symmetries from Offline Robot Interaction Data

Symmetry-aware robot policy learning requires coupled observation-action transformations. For pixel observations, even simple physical symmetries can induce complex, viewpoint-dependent image transformations that are difficult to specify by hand. Existing symmetry-aware methods often rely on representations with predefined transformations, requiring preprocessing tailored to the sensing setup and task. We introduce MIRROR, a framework that learns how a known symmetry group operates jointly on observations and actions in partially observable Markov decision processes (POMDPs) from offline interaction data. MIRROR couples these transformations through adversarial tuple matching over one-step transition tuples, while explicit group-consistency regularization penalizes violations of the group-operation axioms. On four Robosuite tabletop manipulation tasks, MIRROR achieves the lowest aggregate error on three of four symmetry-quality metrics, including a 61% reduction in compatibility error (which measures consistency under transformation composition) relative to an adapted LaLiGAN baseline. Using the learned transformations in frame-averaged behavior cloning (BC) yields higher mean success rates than vanilla BC on all four tasks, increasing average success from 80.48% to 85.12%, with per-task gains up to 9.66 percentage points. We also repeat symmetry discovery and policy training in simulation with heightmap and segmentation-mask observations, then deploy the resulting policy zero-shot on a real Kinova GEN3 arm. Project page: https://pomdp-mirror.github.io.
Oct 12, 2026

DEFlow: Deep Equilibrium Flow Maximum-Entropy Reinforcement Learning

Expressive flow policies can represent multiple high-value actions, but sampling them with few integration steps while retaining tractable action densities remains challenging. We introduce DEFlow, a deep equilibrium flow policy that replaces explicit Euler sampling with implicit Euler integration and solves all intermediate actions jointly as an equilibrium system. We combine implicit differentiation with a discrete change-of-variables formula for the final-action density, enabling off-policy training within soft actor-critic without backpropagating through solver iterations. DEFlow achieves an average relative improvement of approximately 10% in mean successful coverage over the strongest baseline across four Multi-Goal docking variants. It also achieves the highest mean time-averaged return on four of five manipulation tasks, while remaining broadly competitive across five MuJoCo tasks. These results support implicit flow policies for tasks requiring both diversity and precision, at the cost of additional nonlinear and linear solves. Primary
Oct 12, 2026

Constrained Differentiable Policy Learning for Safe Quadrotor Navigation

Differentiable simulation enables efficient policy optimization through analytic gradients, yet safety is commonly incorporated as a weighted penalty, requiring manual tuning of the trade-off between task performance and constraint satisfaction. We instead formulate obstacle avoidance as an explicit constraint and train quadrotor policies using primal--dual gradient descent--ascent through a differentiable simulator. We introduce two key components to enable effective constrained training. First, we propose a bounded clearance-based constraint function that provides informative gradients around the safety boundary while preventing large violations from destabilizing primal--dual optimization. Second, we introduce log-mean-exp (LME) aggregation to smoothly emphasize critical per-step safety violations across a trajectory and training batch within a single constraint governed by one dual variable. Together, these components enable the desired safety level to be specified directly while balancing constraint satisfaction and task completion. We theoretically prove that the proposed optimization with differentiable simulation converges to feasible policy under mild assumptions. Experiments demonstrate improved safety--performance trade-offs over fixed-penalty and existing policy-learning baselines, as well as generalization to denser out-of-distribution environments. Finally, the learned policy transfers zero-shot to Gazebo/PX4 and a physical quadrotor without additional fine-tuning.
Oct 11, 2026

LERR-Nav: Learning to Coordinate Exploration, Recovery, and Reminiscing for Object Navigation

Object Navigation in unseen environments remains a fundamental challenge in embodied AI, as it requires agents to generalize visual and spatial reasoning to novel layouts. While recent advances in large language model (LLM) and vision-language model (VLM) have enhanced semantic mapping, spatial memory, and waypoint selecting for navigation, successful navigation requires more than selecting the next promising waypoint. A robot must also recognize stalled progress, escape unproductive behaviors, and revisit overlooked target cues. Addressing these challenges requires both specialized navigation strategies and a high-level policy that coordinates their execution. Motivated by the analysis of failed navigation episodes, we propose LERR-Nav, an object navigation framework that coordinates exploration, recovery, and reminiscing through a learned high-level policy. The framework comprises four complementary navigation strategies: value-function-guided exploration, LLM-guided exploration, recovery from stalled or repetitive behavior, and memory-guided reminiscing. To support history-aware coordination, we maintain the execution memory that preserves textual summaries of all strategy executions alongside a sliding window of map sequences. To learn the coordination policy, we initialize a high-level vision-language agent from Qwen2.5-VL-7B-Instruct and train it through a two-stage curriculum. The first stage uses map-centric supervised fine-tuning (SFT) to establish spatial grounding. The second stage performs SFT on variable-length navigation segments to learn which navigation strategy to invoke next based on current maps and the hybrid execution memory. Experiments on HM3D and MP3D demonstrate state-of-the-art navigation performance, while ablation studies validate the contributions of individual components.
Oct 10, 2026

VbotNav-0: Asynchronous Cloud-Edge Dual-Policy Navigation across Indoor and Outdoor Environments

Continuous indoor-outdoor navigation connects destination-level reasoning with motion decisions that must respond to current observations. We present VbotNav-0, a cloud-edge dual-policy framework that links these processes through asynchronous policy-to-policy conditioning. The cloud policy produces subgoal guidance from visual history, navigation instructions, and route tokens; the edge policy combines available guidance with current RGB and LiDAR observations to generate local trajectories. This formulation allows the local perception-action loop to continue between cloud updates. A common route representation connects indoor paths and outdoor map routes, and the training recipe combines navigation demonstrations, corrective data, and grounded visual question answering (VQA). In the reported closed-loop evaluation of 258 episodes across 145 scenes, VbotNav-0 achieves 55.04% success rate and 55.03% success weighted by path length.
Oct 9, 2026

Chronos Policy: Learning Robot Policies on Physical Time

Robot policies often convert sensing and control into fixed-rate observation-action sequences, leaving physical time implicit in sequence position. This assumption breaks when observation rates or delays vary, or when inference latency changes. We introduce Chronos Policy, which makes physical time explicit by conditioning on observation capture times and desired action execution times. We study two ways to incorporate time into a diffusion transformer: adding timestamp embeddings to token features, or encoding physical time in positional encoding through time-aware rotary positional encodings (time-aware RoPE). We find that time-aware RoPE is more robust, allowing attention to depend directly on physical time while keeping token content unchanged. During training, we temporally randomize which observations are sampled and when the action chunk begins, exposing the policy to a wider range of sensing and execution schedules that may occur during deployments. This also enables learning from asynchronous, multi-rate data without forcing all measurements onto a fixed temporal grid. Across simulation and real-world experiments, Chronos Policy improves performance on dynamic tasks, where observation-to-action timing is critical, and on multisensory tasks, where the relative timing between observations matters. On standard quasi-static tasks, it preserves nominal performance while improving robustness to changes in sensing and inference timing.
Oct 9, 2026

ATHLETE: Learning Reactive Humanoid Tennis via Trajectory-Guided Motion Matching

Humanoid tennis requires reactive whole-body coordination under rapidly changing ball dynamics, yet collecting task-specific motion-capture demonstrations that cover diverse approach states, contact locations, and timing conditions is costly. Inspired by how humans acquire new skills by combining reusable motor experience with task-specific demonstrations, we present ATHLETE, a framework for learning reactive humanoid tennis from generic locomotion data and ordinary tennis videos. ATHLETE combines LAFAN1 locomotion with video-derived striking motions and uses Trajectory-Guided Motion Matching (TGMM) to construct diverse task-conditioned whole-body references. These references are transformed into a reactive policy through physics-based teacher--student learning with online DAgger distillation, followed by PPO fine-tuning for the tennis task. The resulting policy responds directly to robot and ball observations without requiring reference-motion inputs at deployment. On a real humanoid robot, ATHLETE achieves forehand and backhand hit rates of 90.00%90.00\% and 86.67%86.67\%, respectively, and enables multi-shot rallies with human players. These results demonstrate that ATHLETE can learn dynamic humanoid tennis skills from reusable locomotion data and video demonstrations. Result videos and code can be found at https://anonymous.4open.science/w/ATHLETE2027/.
Oct 9, 2026

Make π0.5 Learn to Call Astra for Reliable Robot Manipulation

GPT-6 Astra provides robust embodied reasoning, but frequent external queries increase inference time and token costs. We equip a frozen π0.5π_{0.5} policy with a lightweight CALL head that learns to predict when assistance is necessary. CALL reuses features from the standard policy forward pass and operates locally without consulting external models. Astra is activated solely for critical recovery events and returns control to π0.5π_{0.5} upon stabilization. Evaluated across 40 LIBERO tasks (200 episodes), our system achieves a 97.5% success rate, outperforming the 95.5% baseline of pure π0.5π_{0.5}. In the Object, Goal, and Spatial suites, Astra executes only 5.49% of environment steps, consuming approximately 5,027 tokens per test episode. This selective intervention preserves the fast ordinary execution of π0.5π_{0.5} while seamlessly integrating Astra's reasoning on demand. Our project page is available at https://kykgeek.github.io/Pi05_CALL_Astra/.
Oct 9, 2026

ReAT: Residual Action Tokenization for Autoregressive Robot Policies

Autoregressive robot policies rely on action tokenizers to convert continuous action chunks into compact sequences of discrete prediction targets. Existing tokenizers can reconstruct action chunks from discrete codes, but often leave implicit what each individual token contributes to the reconstructed action. We introduce Residual Action Tokenization (ReAT), which gives each token an explicit role in refining the current action reconstruction. It uses the decoded reconstruction as feedback to construct the next token from the remaining action error. This closes the loop between encoding and decoding, making each new token responsive to the errors left by the actual decoded prefix. Across 19 manipulation tasks spanning three simulation benchmarks and real-world settings, 8-token ReAT policies outperform the evaluated baselines in task success; on LIBERO, even 4-token ReAT surpasses the evaluated 8-token action-tokenization baselines at half the token budget. These results suggest a path toward action tokenization that supports self-corrective policy learning, helping autoregressive policies learn to progressively refine their own action plans.
Oct 9, 2026

JoyInside-VLM-Agent: Product-Conditioned Multimodal Policy Learning for Embodied Agents

Embodied agents must translate multimodal observations and user requests into executable actions while adapting to product constraints, changing goals, and execution feedback. We present JoyInside-VLM-Agent, a ReAct-style multimodal agent developed through product-conditioned trajectory construction and staged policy learning for robotic platforms and smart hardware. A closed-loop sandbox generates multimodal interaction trajectories covering goal revisions and failure recovery, which are checked against execution outcomes and semantic criteria. Supervised fine-tuning and multi-teacher on-policy distillation precede reinforcement learning with DPER (Dual-Source Privileged On-Policy Self-Distillation with Token-Level Evidence Routing). DPER routes token-level evidence between a frozen privileged teacher and a privileged student sharing the rollout policy's parameters, modulating reward-derived token advantages while preserving their signs. Training rewards combine generative process evaluation, multi-judge rankings, and task-queue consistency. JoyInside-VLM-Agent achieves a BFCL v4 overall score of 77.44, exceeding the evaluated baselines; results on the other public benchmarks are mixed. On JoyInside-Devices and JoyInside-Robotics, the evaluated system, which includes domain-specific skill refinement, leads the baselines on all reported metrics. Two real-world quadruped case studies illustrate visually grounded search, replanning after execution failure, and explicit failure reporting.
Oct 9, 2026

LeapBot-JAC: Joint Action Chains for World Action Models

Action-chunking policies predict future controls but execute only a prefix before replanning. How should a fixed action-token budget be organized to support execution? We propose LeapBot-JAC, which allocates sparse future-control references and dense execution targets within one shared action expert. Both chains are supervised by demonstrated controls and jointly denoised at aligned physical-time positions. Directed attention allows execution tokens to read reference tokens subject to the host's visibility constraints. JAC adds neither action tokens nor parameters and retains the host's world-modeling process; both chains participate in inference, while only the execution chain drives the robot. On LIBERO, JAC reaches 98.3% success with FastWAM and LeapBot-WA, up from 97.0% and 97.3%. On zero-shot LIBERO-Plus, their success rates improve from 53.5% to 70.4% and from 73.1% to 77.3%, respectively. RoboTwin and two real-world tasks provide additional deployment evaluations. Single-training-seed comparisons and FastWAM ablations support action-role organization as a practical design choice under a fixed action-token budget.
Oct 9, 2026

TAME-Nav: Temporal Asymmetric Mixture-of-Experts Distillation for Agile and Versatile Robot Navigation

Agile and reliable navigation is a core capability for embodied robots navigating complex spatial constraints and dynamic interactions. However, a single policy often struggles to accommodate the distinct decision-making demands of diverse scenarios. We present TAME-Nav, an asymmetric Teacher-Student navigation framework built around a temporal soft mixture of experts (Soft-MoE). The Teacher combines clean observations with causally maintained dynamic-obstacle tracks to construct privileged representations, while the Student learns complementary navigation features from corrupted observation histories through temporal encoding, soft expert routing, and latent distillation. Procedural scenario generation and parallel task-level simulation broaden training coverage, while deployment-informed asymmetric domain randomization supports sim-to-real transfer. Across in-distribution and unseen compositional scenarios, TAME-Nav achieves higher success rates, lower collision rates, and higher navigation scores than representative baselines. On a real quadruped robot, it completes all twenty trials across local-trap escape and narrow-passage navigation with sudden dynamic obstacles, with only one minor contact.
Oct 9, 2026

Evidence-Guided Counterfactual Learning for Long-Tailed Robotic Policies

In the practical deployment of vision--language--action policies, handling long-tailed task distributions is a common challenge due to the highly imbalanced composition of robot demonstration datasets. To investigate this issue, we study the head-task action bias in shared robot policies, particularly focusing on its relationship with task-specific visual evidence. Our experimental analysis shows that, when the visual evidence required by a tail task is insufficient, the policy increasingly falls back on action patterns reinforced by data-rich head tasks, resulting in degraded performance on under-represented tasks. Motivated by this observation, we propose Evidence-Guided Counterfactual Learning, a training framework for long-tailed robot policies. Our method identifies decision-relevant visual evidence using instruction-related and action-related signals, and constructs counterfactual observations by removing the selected evidence. The resulting counterfactual predictions are used to strengthen the association among task instructions, visual evidence, and actions, with the supervision adaptively adjusted according to the number of demonstrations available for each task. The proposed framework requires neither additional training data nor region-level annotations and incurs no additional computation during deployment. Extensive experiments across simulated benchmarks, different VLA architectures, and real-world robot platforms demonstrate that our method robustly and consistently improves long-tailed robot policy learning, particularly for tail tasks, while maintaining reliable performance on head tasks.
Oct 8, 2026

Embodied Turing Machines: Stateful Code for Robot Recursive Self-Improvement

Most robot policies keep a model in the control loop: a VLA maps observations to actions, and an Agent Harness, such as Agent-as-Policy or Harness VLA queries a VLM for decision making at run time. We propose a different view: the embodied world is an Embodied Turing Machine, whose tape is the robot and environment state and rules are the policy. If this state can be represented accurately, the decision making can be written entirely in code. We therefore propose Code-Only-as-Policy (COAP): code measures and tracks the robot, environment, and task state from camera images and proprioception, and makes every decision from it. The same code applies across episodes, and different tasks share one library without a VLM or VLA in the loop. Compared with VLAs and Agent Harnesses, we analyze three advantages of COAP: (i) Explicit State: the state can be stored in code; (ii) Execution: code makes decision making controllable, recovers from failures flexibly, and runs fast and cheaply online; (iii) Extensibility: new tasks reuse, inherit, or extend the shared library, so capabilities can accumulate over tasks. These advantages make COAP a suitable medium for recursive self-improvement (RSI): coding agents develop the library in a closed loop, and each change is explicit and controllable. On RoboDojo's 42 bimanual tasks, the resulting library reaches a success rate of 70.24% without a model at test time. The upper bound of COAP lies in how accurately the state is represented for decision making and how robust the code logic is. We thus propose COAP as a new paradigm for embodied tasks; since it applies across episodes, it can also serve as an efficient data engine for VLAs and Agent Harnesses.
Oct 8, 2026

Real-Time Motion Planning with Dynamic Hazards: Classical vs. Learning-Based Methods

We study real-time motion planning in dynamic hazard fields through a controlled comparison between classical planning and learning-based methods. Rather than introducing a new planner, we construct a unified benchmark in which representative classical and learning-based methods face the same environments, motion constraints, information assumptions, and evaluation metrics. The test environment consists of planar domains populated with rotating sprinkler-like hazards that generate time-varying forbidden regions via sweeping angular sectors. Our results show a clear regime shift. In deterministic environments, classical planners achieve near-perfect success and higher-quality paths, though sometimes at the cost of substantial planning or replanning time. Under stochastic obstacle dynamics, however, online search becomes strongly budget-sensitive: low budgets lead to frequent failure, while high budgets improve success at the cost of latency and longer trajectories. PPO-based policies, trained under the same scenario distribution, consistently outperform in latency, success rate, and path quality in these stochastic regimes. Overall, the results indicate that uncertainty in obstacle evolution, more than partial observability, is the dominant factor determining which planning paradigm is practically effective for the problem at hand.
Oct 8, 2026

Residual Modeling Closes the Regression and Generative Policy Gap in Robot Learning

Learning from demonstration has enabled impressive robot behaviors. A common choice for policy learning is to use diffusion or flow matching (Flow-Policies), which often outperforms direct action regression trained with mean squared error (MSE-Policies). This gap is commonly attributed to multimodal demonstrations. We revisit this gap from the perspective of statistical modeling: how action-prediction residuals shape policy optimization. Our analysis of real-world robot demonstration data reveals substantial state-dependent variation in residual scales and heavier-than-Gaussian tails. While both MSE-Policies and Flow-Policies exhibit heavy-tailed action residuals, their training gradients behave differently: MSE allocates more gradient magnitude to observations with large action residuals, which hurts optimization. Motivated by these findings, we introduce heteroscedastic Student-t action regression (HT-Policies), which learns input-dependent residual scales and reduces the influence of heavy tails. HT-Policies predict action chunks with a single feed-forward pass and can reuse pretrained flow-matching-based policy networks as the backbone. Across four simulation benchmarks and real-robot evaluations, HT-Policies achieves success rates competitive with generative policy baselines, both when trained from scratch and from pretrained vision-language-action and world-action models, despite being faster in training and inference. Together, these findings shed light on the practical advantages of generative objectives in robot learning from demonstrations and offer an efficient direct-regression alternative for a range of architectures and tasks. Project page: https://projects.thelabone.org/regression-policy/
Oct 8, 2026

PathTime-VLA: Path-Time Decoupling for Factorized Post-Training of Vision-Language-Action Policies

Vision-Language-Action (VLA) policies typically predict actions at fixed time intervals, coupling the route a robot follows with its execution pace. This coupling complicates adaptation from teleoperation: useful geometric guidance comes with timing shaped by interface delays and operator behavior. Our key insight is to bring the path-time parameterization of classical motion planning into the learned action representation of a VLA. We introduce PathTime-VLA, which represents motion as a progress-indexed interaction path X(s)X(s) and a positive interval-time profile. The latter defines a monotone time law t(s)t(s), yielding controller commands X(s(t))X(s(t)). For a given path, alternative executions are expressed through the time profile, allowing chunk-wise speed choices without changing the geometric prediction target. This representation supports a staged post-training procedure: demonstrations and DAgger interventions establish a target-domain prior, Speed-DQN learns execution multipliers from robot interaction, and Path-AWR uses rollout outcomes to refine the diffusion path generator. A path-conditioned action expert realizes the resulting motions while maintaining distinct learning interfaces for path generation and execution timing. Across three tasks, the complete method achieves 58/6058/60 successes versus 57/6057/60 for PathTime-VLA under BC + DAgger at fixed 1×1\times, with approximately 3939-52%52\% shorter mean completion times over successful trials.
Oct 8, 2026

RoboAware: Learning to Coordinate Embodied Skills from Counterfactual Outcomes

Embodied coding agents can combine modular robot skills with frozen end-to-end policies, yet effective composition requires anticipating which policy family will succeed in the current physical state. We present RoboAware, which builds on coding agents' skill orchestration by learning only a state-conditioned responsibility coordinator from counterfactual outcomes. Inspired by the success of REPL, we propose the P5P^5 schema and formulate a hierarchical MDP based on it. P5P^5 organizes skills uniformly into five semantic stages, defining where responsibility can be compared. To address the lack of counterfactual branch outcomes in existing work, we introduce State-Locked Counterfactual Branching (SCB), which restores the same training state to generate and execute a code block from each admissible family, exposing outcomes that selected-branch experience leaves unobserved. Building on this, we propose Execution-Aware Learning (EAL), which combines Monte Carlo tree search with Q-learning to distill these outcomes into family-conditioned values. At deployment, the coordinator selects the policy family according to observable context, and the frozen coding agent generates the next local code block. Comprehensive single-episode evaluations on 100 tasks show that RoboAware reaches a 77.0% overall success rate, with SOTA averages of 90.0% on RoboSuite, 73.8% on diverse LIBERO-Pro task clusters, and 90.0% on challenging RoboTwin bimanual tasks, outperforming existing code-as-policy and VLA-harness baselines.
Oct 8, 2026

PMTRM: Pseudo-Memory Temporal Re-encoding Module for Embodied Policy Learning

Robotic manipulation often contains repeated motions whose local observations look similar at different phases. When these phases require different actions, a policy that relies mainly on the current observation may repeat completed motions or switch phases at the wrong time. To address this phase ambiguity, we present the Pseudo-Memory Temporal Re-encoding Module (PMTRM), a lightweight plug-in module with only 7.61M parameters that encodes a bounded history of executed states and actions into a latent sequence for existing policies. To help distinguish phases, a temporal heterogeneity objective penalizes positive similarity between distant positions in this sequence, while anchor and reconstruction losses preserve information needed for action prediction. The reconstruction decoder is used only during training, leaving the temporal re-encoder to supply history to the policy at inference. We train the module progressively on synthetic sequences and robot data, then jointly with the policy, using temporal masking to accommodate partial histories. This integration retains the original action head and action space and adds auxiliary losses to the original policy loss. Experiments with multiple policy backbones in simulation and on a real robot show improved task success on tasks with phase ambiguity, with little additional computation.
Oct 7, 2026

Agentic RSR: Real-to-Sim-to-Real through Scene Reconstruction and Execution-Grounded Robot Policies

A simulation of a real robot workspace must preserve task-relevant interactions, while policies developed in it must operate on observations available to the real robot. Yet scene reconstruction and policy development are often treated separately. We present Agentic Real-to-Sim-to-Real (Agentic RSR), a framework that links scene reconstruction, policy development, and real-robot execution through the same manipulation task. Given a workspace video, a task description, and a known robot model, an agent recovers metric scale, iteratively refines the scene using visual feedback, and checks task-relevant interactions in MuJoCo. A coding agent then develops an executable policy, progressing from privileged object poses to visual observations and randomized simulation. The policy can interleave multiple observations and actions within one invocation, while the agent uses execution feedback to continue, retry, or revise its approach. A shared task-level interface carries the policy and accumulated experience to the real robot, where fresh observations and safety checks guide execution. Using Claude Opus 5.5 through DeepSeek Harness (DSH), all 18 scenes across two robots pass the joint visual check; mean four-view Depth MAE against reference depth estimates is 0.0808 m, mean Lab ΔE76ΔE_{76} is 8.01, and mean grayscale SSIM is 0.7338. In real-robot experiments, the aggregate task success rate reaches 80% of the simulation task success rate, indicating substantial retention of simulated performance on hardware. Code and reconstructed scene data will be made publicly available.
Oct 7, 2026

RFPO: Rectified Flow Policy Optimization for Embodied Control

Flow-based policies provide an expressive framework for continuous robot control, but their iterative ODE integration incurs substantial inference cost. Naively reducing the integration budget can severely degrade control, since policies optimized under full-step execution are not explicitly constrained to remain reliable under coarse numerical integration. We refer to this mismatch as the few-step discretization gap. To address this problem, we introduce RFPO, a flow-policy optimization framework for reliable few-step execution. Reward-aware online Reflow rectifies student-induced transport paths during on-policy learning, making the resulting policy more robust to coarse integration. A frozen Gaussian PPO controller supplies complementary action-space supervision at full and intermediate integration budgets, while the deployed policy remains a single flow student executed with one Euler step. Across Unitree Go2, Boston Dynamics Spot, Unitree H1, and Unitree G1, RFPO consistently preserves full-step control performance under one-step execution, with one-step returns remaining within 2.4% of their corresponding 64-step values across both zero and random initialization. On Unitree Go2, one-step execution retains 98.5% of the 64-step reward while reducing onboard mean inference latency from 4.39 ms to 0.08 ms, yielding a 54.9x speedup. Real-robot experiments further validate stable one-step locomotion. Code: https://github.com/AIGeeksGroup/RFPO. Website: https://aigeeksgroup.github.io/RFPO.
Oct 6, 2026

CAP: Codebook-Aligned Prediction for Tokenized Robot Policies

Action tokenization converts continuous robot actions into discrete symbols that can be modeled autoregressively. However, existing tokenizer-based policies typically ignore the tokenizer's learned latent code structure: after tokenization, the policy treats tokens as unrelated class indices and learns a new classifier from scratch. We show that this discarded structure is valuable. We introduce Codebook-Aligned Prediction (CAP), a method that directly reuses the tokenizer's code vectors as policy class prototypes while leaving the tokenizer and policy backbone otherwise unchanged. Across four quantizer families, three simulation benchmarks, and two real-robot tasks, CAP consistently improves task success over standard token classification heads while holding the tokenizer (and therefore its reconstruction quality) fixed. Our analysis further shows that these gains are not explained by higher token accuracy or changes in the policy head alone. Instead, reusing the tokenizer codebook provides the policy with valuable information about the tokenizer's learned latent structure across tokens, making token prediction errors more benign in action space and improving the representations learned by the policy backbone. These results suggest that action tokenizers learn useful action-aware latent structure beyond discrete targets that should be preserved when training downstream policies.
Oct 6, 2026

Beyond Reconstruction: What Matters in Action Tokenization for Robot Policies?

Autoregressive action-token policies such as vision-language-action models require action tokenizers to translate discrete token sequences into precise control actions in continuous space. Many action tokenizers learn the mapping between tokens and actions via a reconstruction objective. However, as we show through extensive analysis, sufficiently accurate action reconstruction is only one part of what makes a downstream robot policy successful. It is also critical that the policy is able to predict the right tokens for new observations, and that unseen policy token predictions still decode into reasonable actions. These properties are downstream of tokenizer training and are not directly incentivized by a reconstruction objective alone. In this work, we introduce Predictable and Robust Action Tokenization (ProAct), a tokenizer training method that strategically augments reconstruction with the goal of improving downstream predictability and robustness. ProAct is policy-agnostic and uses only action datasets for training. Across the Robomimic, LIBERO, and RoboTwin benchmarks and a diverse set of tokenizer architectures, ProAct improves rollout success by an average of 11.3 percentage points. These improvements also translate to vision-language-action policies and real-world robotic manipulation, yielding average gains of 21.8 and 36.7 percentage points, respectively. These results suggest that effective action tokenization should be designed as a policy interface that balances fidelity, predictability, and robustness, rather than as a reconstruction problem alone.
Oct 6, 2026

RIWANav: Recursive World-Action Models with Self-Improvement for Urban Navigation

Long-horizon urban navigation requires sequential local decisions whose errors can compound over time. Imitation learning (IL) rarely learns from failures, while physical trial-and-error reinforcement learning (RL) is costly. Action-conditioned world models can provide imagined feedback by predicting visual consequences for candidate actions. However, a frozen world model may become less reliable as the policy evolves. In this paper, we introduce RIWANAV, a post-training framework that casts the coupled adaptation of a world model and an action model (policy) as task-specific recursive self-improvement (RSI). Each cycle alternates two updates. The world model evaluates policy actions through imagined outcomes, providing comparative feedback for group-relative policy optimization (GRPO). The improved policy then constructs a grounded self-curriculum, selecting expert-consistent action-video pairs by behavioral novelty and prediction error. The refined world model supplies feedback for the next policy update, closing the recursive self-improvement loop. Experiments show that RIWANAV outperforms training baselines and prior methods, validating the proposed recursive self-improvement loop between the policy and world model. Real-world trials further demonstrate its practical applicability.
Oct 6, 2026

Towards Efficient Robotic Manipulation Models with Self-Recursive Pruning

Network pruning can reduce parameter redundancy in robotic policies. However, generic pruning criteria are tailored for image recognition tasks and commonly designed to preserve weight magnitude, local reconstruction, or language-model likelihood rather than closed-loop action behavior. Directly applying these pruning algorithms to robotic tasks yields unsatisfactory performance. In this paper, we propose Loss-Conditioned Activation-Moment (LCAM) pruning, a training-free method for unstructured pruning of pre-trained robotic manipulation policies. Specifically, we first rank connections using row-normalized weight contribution, activation moments measured on calibration demonstrations, and the sensitivity of output directions to the action-prediction loss. We further design a self-recursive coarse-to-fine procedure: importance is recalibrated after each nested coarse pruning stage, while held-out offline action distortion guides fine-grained budget allocation after a sparsity knee. Our algorithm is free from costly recovery training and simulator rollouts after pruning. Experiments on three LIBERO suites with competitive robotic policies, together with evaluations on OpenVLA, show that LCAM attains competitive performance across a broad range of pruning ratios. Notably, on LIBERO-Object with OpenVLA, our LCAM achieves 84.0% success at 50% unstructured pruning, retaining over 90% of the dense policy's success rate. Promising results on real-world robotic ping pong further demonstrate the effectiveness of our pruning algorithm.
Oct 6, 2026

Compact Robot Policies Need Fine-Grained Visual Representations

Multi-task manipulation policies differ in architecture, scale, and pretrained priors all at once, so published comparisons cannot attribute performance to any single component. We argue that most of it comes from the visual representation, and that parameter scale and generative priors are largely incidental. To test this, we build CoRP (Compressed Representation Policy), a deliberately compact policy (48.9M parameters, no vision-language model and no video-generative prior) that factorizes into a representation extractor and a flow-matching action generator. It reaches 97.0% on LIBERO and 75.78%/73.36% on RoboTwin 2.0 Clean/Randomized, matching systems 40.9-163.6x larger. Holding the action generator fixed, we then vary one extractor property at a time. Pretrained initialization is decisive: a random ViT-S/14 drops to 78.1% and an ImageNet ResNet-34 to 74.5% on LIBERO. Pretraining alone is not enough, as freezing the encoder costs 19.8 points. Compression matters as much: resampling each view to 48 tokens beats passing all patch tokens (97.0% vs 83.2%), and a variational information bottleneck over those tokens is worse than a hard token budget, cutting LIBERO-Goal from 95.8% to 33.0% by suppressing the instruction-dependent token selection the policy relies on. Language conditioning contributes only where the observation leaves the goal ambiguous (LIBERO-Goal: 9.2% to 95.8%), while on RoboTwin 2.0, where observations are unambiguous, removing it slightly improves success. Therefore, we argue that a compact policy works when its representation is pretrained, task-adapted, and compressed. Project page: https://corp-policy.github.io/
Oct 6, 2026

Modeling Latent Disturbances for Robust Decision-Making in World Models

In this paper, we study robust decision-making in the latent space of world models (WMs). Robust optimization is a mathematical framework where, given explicitly specified dynamics and physically meaningful disturbances, a robot can select actions that remain effective even under worst-case disturbances. However, applying this principle to the learned latent space of WMs introduces a fundamental challenge: because WMs have fully learned state spaces and dynamics inferred from high-dimensional observations, it is unclear how to define latent-space disturbances that faithfully represent uncertainty in the underlying system. Our key idea is to model a latent-space disturbance as a perturbation to the learned latent dynamics that induces pessimistic but plausible transitions. Specifically, we construct a set of plausible latent dynamics by combining a dynamics-aware similarity metric that captures plausible transitions with out-of-distribution detection that excludes implausible latent states. We calibrate this uncertainty set over latent dynamics using conformal prediction, ensuring that WM imaginations induced by the latent disturbance remain plausible without becoming overly pessimistic. We then jointly optimize robust robot actions and the worst-case latent disturbances through game-theoretic optimization. We leverage this latent-space robust optimization to robustify policy steering, considering two paradigms: latent safety filtering and sample-and-verify steering of a generative control policy. Our controlled simulation experiments show that our latent disturbance enables robust decision-making directly in WM latent spaces, and hardware experiments with a Franka manipulator show that modeling latent disturbances enables robust policy steering, reducing failures by 70% in safety filtering and 54% in sampling-based policy steering. Project website: https://junwon.me/LatentDisturbance/.