Robot Policy Adaptation

Behavior-cloned policies often learn multiple behavior modes from demonstration datasets, including modes that are unsafe or otherwise undesired at deployment. For example, a policy trained on diverse handover demonstrations may learn to pass a knife blade-first. Standard remedies such as data curation and inference-time steering either require access to the original demonstrations for full retraining or add substantial inference-time overhead. To address this gap, we propose MoRE(Mode Redirection), which redirects policy rollouts toward desired behavior modes through a short "uncloning" step. Specifically, MoRE distills the redirection signal from a temporary mode classifier into the policy weights to steer behavior. A retain loss balances this edit by preserving desired-mode competence, allowing the standalone policy to suppress unwanted modes with zero inference-time overhead. Across eight simulated and real-world tasks, MoRE improves the average deployment success rate (SR) by 44 percentage points over the original mixed-mode policy. Among all compared adaptation and steering baselines, MoRE achieves the strongest SR and approaches the filtered-data retraining reference, while preserving task competence and inference speed. MoRE also generalizes across robot policy backbones, including Diffusion Policy and the Pi0.5 VLA, diverse task categories, and real-world deployments.

Behavior-cloned diffusion policies are expressive but remain vulnerable to covariate shift: small deviations from demonstrated states can compound into task failure. Existing methods address this either by expanding the training distribution through expert corrections or synthetic augmentation, or by steering a frozen policy at test time with guidance from a learned model. The former can be expensive or assumption-dependent, while the latter discards the corrected trajectories after execution. We introduce ReGuide, a self-improving framework that treats guided rollouts as reusable on-policy recovery data. ReGuide first uses Phase-Conditioned Guidance (PCG) to generate corrective rollouts: it constructs phase-specific latent targets, applies guidance only in the drifted-but-recoverable regime, and guides through the estimated clean action to match the dynamics model's training distribution. Successful guided rollouts are then absorbed back into the policy through ReGuide-FT, which fine-tunes the current checkpoint, or ReGuide-FS, which retrains from scratch on the augmented dataset; the two can also be composed and iterated. On Robomimic Can, Square, Transport, and Tool Hang, ReGuide improves base-policy success by $1.3$--$7.7\times$, outperforms LPB in the test-time-only setting, and matched-data ablations show that the gains come from guided recovery data rather than additional rollouts alone.

Robots trained on real world data tend to be imprecise, slow, and brittle to perturbations. Improving these policies with reinforcement learning (RL) is an appealing alternative, but this process often requires expensive training in the real world. Performing policy improvement in simulation instead provides a far cheaper alternative, but unconstrained RL in simulation can exploit contact and dynamics mismatches, resulting in unsafe behaviors that do not transfer to hardware. Common forms of regularization can furthermore limit improvement by overconstraining to an imperfect behavior prior. In this work, we propose Support-Constrained Off-Domain REinforcement (SCORE), a real-to-sim-to-real framework that constrains RL in simulation to the support of a generative policy pretrained on real data. We instantiate this constraint through flow steering, restricting SCORE to actions the base policy can already produce, which ensures transferable behaviors while maximizing policy improvement. Improving a policy with SCORE requires minimal effort: it learns from sparse rewards, avoids distillation, and leaves the base policy untouched. Across eight real-world dexterous multi-fingered robotic manipulation tasks, SCORE improves average success rate from 37.8% to 89.9%, compared to 59.5% for the best baseline, and reaches success in 36.8% fewer steps than the base policy. Ultimately, through extensive experiments and ablations, we show that simulation can substantially improve real-world manipulation policies when policy optimization is appropriately constrained, introducing a new paradigm for real-to-sim-to-real policy improvement. Videos and code are available at https://weirdlabuw.github.io/score/.

Robots deployed in the real world rarely operate under a single fixed dynamics model: wind changes, payloads vary, batteries drain, contacts shift, and hardware wears. Yet most learning-based controllers are trained once and deployed as if learning were complete. This prevents the robot from using deployment experience to further improve task performance. In this work, we propose a continual learning framework that uses real-world experience to improve robot policies under hidden and recurring dynamics. Our method learns a condition-aware dynamics model from real state-action trajectories by combining an analytical physics prior with a neural residual for unmodeled effects. A recurrent encoder infers the current hidden condition from recent interaction, and this estimate conditions both the residual model and the policy. Policy learning is performed via differentiable simulation using diverse learned dynamics sampled from the latent model. At deployment, these sampled conditions are replaced by conditions inferred online from recent real interaction, allowing the policy to recover recurring dynamics by recognition rather than residual re-fitting. Through extensive simulation studies and real-world experiments, we demonstrate that the framework improves policy performance under diverse unobserved disturbances. On real quadrotor trajectory tracking under changing wind, the policy recovers from recurring disturbances in roughly 1s, about 5x faster than online residual re-fitting. It also reduces large-disturbance hover and tracking errors by 65.7% and 53.3% over the state-of-the-art online adaptation approaches

A central challenge in deploying learned robot policies is inference-time behavior steering: redirecting a policy at test time to satisfy user preferences not anticipated during training, without retraining. Existing methods fail in two modes: end-to-end methods require fine-tuning or expert-level guidance, while neuro-symbolic methods rely on predefined symbols whose edits can result in logically reasonable but physically infeasible plans. To address this challenge, we propose ReStruct, which builds upon a neural automaton policy that decomposes a visuomotor policy into a high-level state-machine skeleton capturing task structure and a low-level continuous controller represented as a residual policy. Specifically, ReStruct adopts the automaton to represent the preference and incorporates it into the skeleton through a synchronous product, thereby reconfiguring the task structure. With the controller kept frozen, the action priors provided by the skeleton are updated accordingly to enable physically-aware control under a modified task structure. Extensive experiments from simulation and real-world show that ReStruct steers a wide range of preferences, from object-centric specifications to temporal-logic constraints, and after steering surpasses existing methods, exceeding VLA models in both task success and preference-following by up to 25%.

Effective online adaptation of vision-language-action (VLA) models remains challenging, as sparse rewards provide weak supervision for high-dimensional autoregressive action policies. Although self-distillation can in principle provide denser training signals, we find that text-based privileged teachers conditioned on demonstrations, retrieved experiences, or high-level plans are ineffective for VLA adaptation, exposing a modality gap between symbolic guidance and low-level robot actions. We propose ROAD-VLA, an advantage-guided self-distillation framework that constructs a proximal teacher directly in action space by perturbing action-token logits with calibrated advantage estimates. This converts sparse rewards into dense token-level supervision while keeping the teacher close to the current policy. We further derive a policy-improvement lower bound under calibrated advantages and accurate teacher matching. Across seven robotic manipulation environments with in-distribution and out-of-distribution shifts, ROADVLA outperforms PPO in nearly all settings, demonstrating robust online VLA adaptation.

In robot learning, scaling training datasets across diverse embodiments and environments has become a dominant paradigm for learning generalizable robot policies. These policies are commonly trained via behavior cloning to imitate actions from pre-collected demonstrations. However, since robot actions are tied to the dynamics of the data collection robot, different robots may require different actions to achieve the same motion. This discrepancy hinders both policy training and deployment across diverse robots. To address this, we propose using Cartesian state delta as a universal action representation across robots, and introduce State Prediction and Adaptive Command Execution (SPACE) framework. SPACE handles robot dynamics variation at three levels: across different embodiments, across hardware units of the same embodiment, and within a single robot during operation. It consists of two components: (i) a Cartesian state delta policy that predicts geometric end-effector displacement, and (ii) Action Adapter, which converts the predicted Cartesian state delta into robot-specific control commands. Experiments show that SPACE substantially outperforms policies that directly predict control commands when learning from data collected across different embodiments and across hardware units of the same embodiment. SPACE also remains robust under dynamics shifts at deployment, including changes in control frequency, object weight, and controller gains. The project page is available at http://haeone.site/space-website/.

A natural recipe for intelligent robotic decision-making is initializing from pretrained generative control policies, which have summarized offline experience, and adapting them to self-collected online experience. We present DF-ExpEnse, an exploration technique that improves the quality of online experience collection, thus increasing finetuning sample-efficiency. DF-ExpEnse leverages the multimodal modeling capabilities of the generative control policy to create an expressive and tractably evaluatable candidate set. It then utilizes an ensemble of critics to identify the action that best balances quality with high exploration interest. In fleet settings, DF-ExpEnse further enables cross-agent communication to facilitate collaborative exploration as a group. DF-ExpEnse can be seamlessly integrated with existing strategies that finetune pretrained generative control policies via reinforcement learning. We experimentally validate consistent sample-efficiency benefits through DF-ExpEnse across a variety of manipulation and locomotion tasks, compared to default finetuning and alternative action selection schemes. Project can be found at https://df-expense.github.io.

Perceptive legged locomotion over discontinuous terrain (e.g., stairs, gaps, and obstacles) requires adaptive behavior, as a single conservative gait cannot produce the anticipatory maneuvers needed for abrupt topology changes. Cast as multi-task reinforcement learning, this problem introduces a tension between sharing and separation. Tasks use a common locomotion base but have conflicting rewards, so a policy must share behavior while avoiding value interference. Prior work addresses only one side, with monolithic policies sacrificing specialization and hierarchical sub-policies sacrificing generalization across transitions and unseen terrain. We propose CTS-MoE, which combines a dense mixture-of-experts actor with perception-based gating to compose shared behaviors and a multi-critic with task-specific value heads to prevent interference. The model is trained end-to-end in a single-stage concurrent teacher-student setup that handles partial observability and avoids sequential distillation, with task labels used only during training. At deployment, routing depends solely on perception, allowing terrain adaptation without a high-level selector or terrain classifier. Experiments on a Unitree Go1 in simulation and on hardware across seen and unseen terrains show task-aware specialization, with lower tracking error and higher success rates than monolithic baselines. Project Website: https://cts-moe.github.io/ .

Robots deployed in the real world should learn from their experience and improve over time. This requires a mechanism of practicing and learning from feedback. In this paper, we propose VERITAS, a generator-verifier framework for generalist robot policies for inference-time policy steering and self-improvement. We use a pre-trained generalist robot policy as a generator'' and pair it with a gradient-free visual verifier'' that evaluates actions at inference time. This framework enables inference-time steering that improves policy performance without additional training. We demonstrate that inference-time verification consistently outperforms vanilla generalists without training on additional demonstration data. Additionally, we demonstrate that the verified rollouts provide effective supervision for offline policy improvement: policies fine-tuned on verified self-generated trajectories achieve consistent performance gains. Notably, we find that post-training with verified rollouts achieves comparable efficiency to expert demonstrations, while requiring no human interventions. Our results highlight inference-time verification as a practical and scalable mechanism for improving robotic policies during deployment.

Robots operating in non-stationary environments must continually adapt their policies as the dynamics drift, but onboard energy and compute budgets cap how often a full state estimation and re-planning step can be performed. This raises a question: \emph{when}, along a horizon, should a robot spend its limited budget? We formulate this problem in time-varying Markov decision processes (TVMDPs) with a known bound on the rate of transition drift. We model execution as a \emph{skip-update} scheme in which, at chosen update times, the agent estimates the transition kernel by maximum likelihood and computes a finite-horizon policy, and between updates reuses this policy under a propagated state estimate. We analyze the dynamic regret of this scheme and show how it grows during skip intervals in terms of the properties of the TVMDP and the skip lengths; the resulting bound answers the opening question via an online, regret-guided update rule that allocates the budget adaptively. We evaluate the rule in a simulated Mars-rover navigation task with time-varying slip dynamics and on a Crazyflie quadrotor in indoor obstacle fields. Adaptive allocation outperforms other budgeted baselines.

Inference-time steering adapts pre-trained generative robot policies during deployment by verifying candidate actions before execution. While prior methods typically perform this verification only with visual observations, vision alone is often insufficient for contact-rich manipulation, where success depends on both global task progress and subtle local interactions such as contact force. We introduce ViTaL, a visuo-tactile inference-time steering framework that formulates multimodal guidance as a bi-level optimization problem. At the high level, visual sampling-and-verification performs long-horizon mode selection, deciding what behavior the robot should execute. At the low level, tactile-guided diffusion editing refines the selected action sequence over a shorter horizon to satisfy local contact requirements. To support outcome-based steering, ViTaL learns a visuo-tactile latent world model and employs semantically aligned visual and tactile verifiers, including a novel text-conditioned tactile reward that scores predicted tactile futures directly in latent space. Across three real-world contact-rich manipulation tasks, ViTaL improves overall success by 51% over the base policy, outperforms unimodal steering by at least 33%, and exceeds naive multimodal fusion by at least 20%. Website: https://yilin-wu98.github.io/vital_website.

Robotic systems that assist humans should be capable of adapting their behaviors to individual user preferences. For instance, users may want a robot arm to adjust the amount of force it applies while folding their laundry or cleaning furniture. Natural language provides an intuitive way for humans to communicate such preferences. Recent progress in language-conditioned robot policies has shown that robots can successfully use language prompts to determine what task to perform. However, extending the same approach to realize how the task should be performed requires detailed labels describing the preferences or styles of trajectories in the task data. Not only is collecting such annotations challenging, but conditioning directly on these labels may also fail to provide fine-grained control over a continuous range of behaviors. For example, it can be difficult to convey the exact force that a robot must apply through abstract instructions like "apply a bit more pressure than before". Therefore, in this work, we propose using language to reason over preferred behaviors instead of directly generating them. We first learn a structured latent representation that organizes user preferences according to differences in the corresponding trajectories. Then, given a preference prompt, we use a foundation model to interpret this latent space and choose a value that produces the desired behavior. Through both simulation and real-world experiments, we show that selecting robot behaviors from an intuitively structured latent space enables more precise adaptation to user preferences while requiring significantly fewer preference labels than language-conditioned policies.

Real-world robots need to adapt their behavior beyond the envelope of their pre-trained policy. Policy finetuning or retraining are options, but they risk catastrophic forgetting, degrading the pretrained policy's base performance. To combat this, we introduce CLAE: Closed-Loop Affine Activation Editing, an inference-time framework for steering the behavior of a frozen policy by editing intermediate activations while keeping the base policy weights and downstream action head untouched. CLAE approaches behavior steering as a closed-loop problem whose outputs edit policy activations that adapt online to the robot state, environment, target behavior, and multi-robot context. It trains a sparse autoencoder over frozen-policy activations, selects behavior-relevant latent features via post-hoc probing, and learns a lightweight RL-based steering policy that applies state-dependent affine edits to selected latents during inference. We validate CLAE on a frozen multi-quadrotor navigation policy trained to perform a single task: navigating robots to a set of goal locations while avoiding obstacles. Through extensive simulations and physical tests, we show that while navigating to their goal positions, CLAE can 1. steer individual robot behavior by controlling each robot's velocity profile; 2. coordinate multirobot behavior by preserving a desired formation; and 3. produce entirely new behavior wherein robots are required to reduce their exposure to surveillance cameras in the environment. Our project website is available at https://clae-robotics.github.io/.

Diffusion policies (DPs) have emerged as expressive policy representations for robot learning, often used with imitation learning methods such as behavioral cloning (BC). However, while their success has largely been confined to BC, direct reinforcement learning (RL) fine-tuning remains challenging because actions are generated through a multi-step denoising process. In this work, we propose MODIP, a framework for the offline-to-online fine-tuning of DPs. Rather than directly applying RL to the DPs, MODIP leverages a world model (WM) to guide policy adaptation and keeps the simplicity and stability of BC. We utilize model predictive control (MPC) to generate high-quality trajectories within the WM, and use them as supervised targets for fine-tuning the DP. To make MPC planning efficient, MODIP uses a terminal state value instead of a policy-dependent state-action value, reducing inference time. Additionally, MODIP trains critics with policy-independent TD targets, reducing training time. Experiments on D4RL (MuJoCo, Kitchen) and RoboMimic tasks show that MODIP improves diffusion policies beyond BC, and is competitive with or outperforms diffusion policy RL fine-tuning methods and strong model-based baselines such as TD-MPC2.

Vision-Language-Action (VLA) models are emerging as a promising paradigm for robotic manipulation, enabling general-purpose policies trained from large corpora of demonstrations and action labels. However, adapting these models to new tasks still typically requires task-specific demonstrations, action annotations, and additional fine-tuning, making deployment costly and difficult to scale. We propose WIZARD, a weight-space meta-learning framework that sidesteps task-specific fine-tuning by generating task-specific LoRA parameters for a frozen VLA policy. Given only a language instruction and a short demonstration video, WIZARD predicts the corresponding adaptation weights in a single forward pass, without target-task action labels or test-time optimization. During meta-training, WIZARD learns to map task evidence directly to expert LoRA updates, capturing relationships between tasks in weight space. Experiments on LIBERO show that WIZARD improves performance by up to ~2x on unseen dataset collections and up to ~14x on unseen tasks. On a Franka Emika Panda, WIZARD consistently improves over a real-domain adapted baseline, showing that generated adapters provide task-level specialization beyond simulation.

Leveraging prior knowledge from pretrained policies, foundation models, or human operators offers an efficient alternative to learning robot skills from scratch. However, these agents often provide actions that are suboptimal, noisy, or misaligned with task-specific expert behavior. We propose GLOVES, a family of flow-based adaptation methods that correct non-expert actions by transporting them toward an expert action distribution. Rather than replacing agentic control with full autonomy, GLOVES performs selective action-level adaptation, improving task success while preserving agent intent. The learned flow also provides a natural in-distribution scoring mechanism through reverse flow evaluation. We use this signal as an intervention gate: actions that appear consistent with the expert distribution are passed through unchanged, while anomalous or out-of-distribution (OOD) actions are corrected. In this way, assistance is only provided when necessary. GLOVES requires only limited expert supervision, using a small number of demonstrations or reusable successful skill segments. By learning local expert action patterns and stitching them during execution, GLOVES provides a lightweight shared-control module for robust action adaptation across tasks and environments. Code and demos are available at ripl.github.io/GLOVES_web.

A policy tuned for one robot often behaves differently on another, whether due to the sim-to-real gap, unknown payloads, or the differing dynamics of two instances of the same robot. In contact-rich, dynamic manipulation, even small motion discrepancies can result in failure to track reference motion, since they disrupt the timing and modes of contact. Common remedies, such as domain randomization or system identification, either produce overly conservative task policies or require data that must be recollected for each robot or payload. We introduce the Torque Adaptation Module (TAM), a learned module that adapts the torque commands sent to the robot to match the behavior of an ideal robot. TAM operates between the low-level controller that tracks the policy's actions and the robot's torque interface. It includes a history encoder that embeds proprioceptive history into a latent state and a torque adaptor that computes residual torque corrections. Because TAM depends only on proprioceptive history and not on policy observations, or the action space, the same TAM weights can be reused to adapt policies with different action spaces (joint targets, end-effector targets, or direct torques). The policies themselves do not need to be trained with domain randomization of robot parameters. Instead, we offload the need for domain randomization to TAM by training it entirely in randomized simulation, using multi-robot pretraining followed by a robot-specific fine-tuning step that still requires no real-robot data. We evaluate TAM zero-shot on a real Franka Panda robot across dynamic manipulation tasks that include a vision-based box pushing policy (from RL), a flip policy (from BC), and an MPC ball-on-plate balancing. Our experiments show that TAM improves zero-shot real-robot execution compared to online system identification and RMA baselines and enables robust dynamic manipulation performance.

Humanoid loco-manipulation requires stable whole-body control under varying object masses and pickup/placement heights. This becomes particularly challenging in sim-to-real transfer, where object-induced load variation and robot-side dynamics mismatch interact during physical contact. Existing history-based adapters often compress these factors into a single latent representation, which can weaken robustness under heavy-load manipulation. We propose \textbf{SplitAdapter: Load-Aware Humanoid Loco-Manipulation via Factorized Adaptation}, which freezes a pretrained box manipulation policy and extends it with object/load and dynamics-aware context encoders trained with split world-model objectives, GRL-based cross-adversarial regularization, and hierarchical Feature-wise Linear Modulation (FiLM). In sim-to-sim experiments and real-world deployment, SplitAdapter improves Full-task success over the base policy and world-model FiLM baselines across object masses of $2$, $4$, and $6$ kg and pickup/placement heights of $0$, $30$, and $60$ cm, with the largest improvements under heavy-load conditions.

Real-world dynamics shifts pose a critical challenge for reinforcement learning in robotics, as policies tightly coupled to nominal environments often fail catastrophically when physical conditions change. Most existing methods rely on encoding explicitly identified physical parameters into a latent context, a parameter-centric paradigm that depends on pre-specified axes of variation and becomes brittle under unmodeled or compound dynamics changes. We revisit dynamics adaptation from an outcome-centric perspective: rather than telling policies what the dynamics are, we enable them to learn how dynamics affect interaction outcomes. Theoretically, this is grounded in a monotonic relationship between target-domain regret and the Lipschitz constant of a trajectory dynamics encoder. Practically, this constant can be upper-bounded through contrastive learning, yielding a smooth, task-relevant latent topology without privileged dynamics information. On MuJoCo benchmarks, our method consistently outperforms parameter-centric baselines under severe dynamics shifts, including unmodeled and time-varying parameters, while also improving in-distribution stability and latent interpretability. Overall, these results validate that controlling latent geometry is a principled mechanism for robust adaptation.

Behavior cloning with high-capacity generative policies achieves strong imitation performance, but is often limited by demonstration coverage and distribution shift. Direct reinforcement learning fine-tuning can improve performance, but updating large action decoders is frequently unstable and sample inefficient. We propose Lagrangian Perturbation Diffusion Steering (LP-DS), a lightweight adaptation method that improves a frozen generative policy by learning a compact noise-space perturbation before decoding. LP-DS optimizes this perturbation with a Lagrangian trust-region objective, improving downstream value while constraining deviation from the latent prior. Across RoboMimic manipulation, OpenAI Gym locomotion, and Adroit dexterous manipulation benchmarks, LP-DS improves sample efficiency, success, and return while maintaining higher action-space entropy than unconstrained noise-space steering, with return improvements of up to 25% over prior baselines. Additional evaluations with flow-matching backbones, a large vision-language-action model, and physical Franka deployment show that LP-DS is not limited to compact diffusion policies or simulated benchmarks. Project page: https://sites.google.com/view/lp-ds/home.

We propose Parameterized Diffusion Policy (PDP), a framework for learning diffusion policies conditioned on low-dimensional, continuous parameters embedded in a learned behavior manifold. By constructing this manifold so that distances between latent representations reflect the semantic similarity between physical trajectories, we transform diffusion from a mechanism for stochastic diversity into a precise and optimizable tool for behavior steering. Our approach enables smooth interpolation between known strategies and efficient adaptation to novel constraints without updating policy weights. We demonstrate that PDP significantly improves adaptation performance on complex multimodal benchmarks in both simulated and real-robot experiments compared to standard diffusion policies, particularly in scenarios requiring the synthesis of novel behaviors.

We present CoRMA(Contrastive Robotic Motor Adaptation), a context-based meta-adaptation framework that modifies RMA for force-dominant assembly. CoRMA replaces raw simulator-parameter adaptation with a compact 6D simulator-only semantic contact context describing contact onset, lateral engagement, guided transition, contact direction, and jamming. A deployable causal Transformer adapter infers this context online from force, proprioceptive, and action histories using semantic regression and a force-regime contrastive objective. At deployment, oracle context is removed and replaced by the inferred context, enabling within-episode adaptation without demonstrations, privileged inputs, or gradient updates. We evaluate CoRMA on PegInsert, GearMesh, and NutThread in Isaac Lab / Isaac Sim 5.0 and on a real Marvin arm. Compared with FORGE baselines that achieve high simulation success but degrade substantially on hardware, CoRMA retains higher verified real success under controlled target-pose noise. These results support semantic contact inference as a reusable adaptation interface within a related assembly task family, while broader unseen-task generalization and Real2Sim calibration remain future work.

Pretrained imitation policies have become a strong foundation for robot manipulation, but they often require online improvement to overcome execution errors, limited dataset coverage, and deployment mismatch. A central question is therefore how reinforcement learning (RL) should adapt policies after offline pretraining. Existing lightweight methods commonly apply residual corrections directly in action space, but this often leads to noisy and poorly structured exploration. In this work, we propose Z-Perturbation Reinforcement Learning (ZPRL), an approach that steers pretrained policies through a compact bottleneck latent rather than through policy weights or output actions. During offline training, we augment the policy with a plug-and-play variational information bottleneck (VIB) module to extract a task-relevant latent interface from observation embeddings. During online finetuning, the base policy is frozen and RL learns only a residual perturbation on this latent, whose decoded representation conditions the frozen action generator. We instantiate ZPRL on flow-matching policies and evaluate it on eight simulation tasks and four real-world tasks. Across diverse manipulation settings, ZPRL improves both sample efficiency and final performance over strong post-training baselines. In the real world, ZPRL improves the average success rate on four tasks by 33.7% over imitation base policies while producing smoother exploration behaviors than an action residual counterpart. These results suggest that a compact, task-aligned bottleneck latent provides an effective interface for online RL adaptation. More videos can be found at https://manutdmoon.github.io/ZPRL/.

Fine-tuning pre-trained robot policies with reinforcement learning (RL) often inherits the bottlenecks introduced by pre-training with behavioral cloning (BC), which produces narrow action distributions that lack the coverage necessary for downstream exploration. We present a unified framework that enables the exploration necessary to enable efficient robot policy finetuning by bridging BC pre-training and RL fine-tuning. Our pre-training method, Context-Smoothed Pre-training (CSP), injects forward-diffusion noise into policy inputs, creating a continuum between precise imitation and broad action coverage. We then fine-tune pre-trained policies via Timestep-Modulated Reinforcement Learning (TMRL), which trains the agent to dynamically adjust this conditioning during fine-tuning by modulating the diffusion timestep, granting explicit control over exploration. Integrating seamlessly with arbitrary policy inputs, e.g., states, 3D point clouds, or image-based VLA policies, we show that TMRL improves RL fine-tuning sample efficiency. Notably, TMRL enables successful real-world fine-tuning on complex manipulation tasks in under one hour. Videos and code available at https://weirdlabuw.github.io/tmrl/.

Large-scale pretraining has made Vision-Language-Action (VLA) models promising foundations for generalist robot manipulation, yet adapting them to downstream tasks remains necessary. However, the common practice of full fine-tuning treats pretraining as initialization and can shift broad priors toward narrow training-distribution patterns. We propose PriorVLA, a novel framework that preserves pretrained priors and learns to leverage them for effective adaptation. PriorVLA keeps a frozen Prior Expert as a read-only prior source and trains an Adaptation Expert for downstream specialization. Expert Queries capture scene priors from the pretrained VLM and motor priors from the Prior Expert, integrating both into the Adaptation Expert to guide adaptation. Together, PriorVLA updates only 25% of the parameters updated by full fine-tuning. Across RoboTwin 2.0, LIBERO, and real-world tasks, PriorVLA achieves stronger overall performance than full fine-tuning and state-of-the-art VLA baselines, with the largest gains under out-of-distribution (OOD) and few-shot settings. PriorVLA improves over pi0.5 by 11 points on RoboTwin 2.0-Hard and achieves 99.1% average success on LIBERO. Across eight real-world tasks and two embodiments, PriorVLA reaches 81% in-distribution (ID) and 57% OOD success with standard data. With only 10 demonstrations per task, PriorVLA reaches 48% ID and 32% OOD success, surpassing pi0.5 by 24 and 22 points, respectively.

Diffusion-based vision-language-action (VLA) models have emerged as strong priors for robotic manipulation, yet adapting them to real-world distributions remains challenging. In particular, on-robot reinforcement learning (RL) is expensive and time-consuming, so effective adaptation depends on efficient policy improvement within a limited budget of real-world interactions. Noise-space RL lowers the cost by keeping the pretrained VLA fixed as a denoising generator while updating only a lightweight actor that predicts the noise. However, its performance is still limited due to inefficient autonomous exploration. Human corrective interventions can reduce this exploration burden, but they are naturally provided in action space, whereas noise-space finetuning requires supervision over noise variables. To address these challenges, we propose UniSteer, a Unified Noise Steering framework that combines human corrective guidance with noise-space RL through approximate action-to-noise inversion. Given a human corrective action, UniSteer inverts the frozen flow-matching decoder to recover a noise target, which provides supervised guidance for the same noise actor that is simultaneously optimized via reinforcement learning. Real-world experiments on diverse manipulation tasks show that UniSteer adapts more efficiently than strong noise-space RL and action-space human-in-the-loop baselines, improving the success rate from 20% to 90% in 66 minutes on average across four real-world adaptation tasks.

Inference-time guidance is essential for steering generative robot policies toward dynamic objectives without retraining, yet existing methods are largely confined to chunk-based architectures that exhibit high latency and lack the reactivity needed for test-time preference alignment or obstacle avoidance. In this work, we formally derive the optimal guidance term for Stochastic Interpolants (SI) by analyzing the value function's time evolution via the Backward Kolmogorov Equation, establishing a modified drift that theoretically guarantees sampling from a target distribution. We apply this framework to real-time control through the Streaming Stochastic Interpolant Policy (SSIP), which generalizes the deterministic Streaming Flow Policy (SFP). Unifying this guidance law with the streaming architecture enables fast and reactive control. To support diverse deployment needs, we propose two complementary mechanisms: training-free Stochastic Trajectory Ensemble Guidance (STEG) that computes gradients on-the-fly for zero-shot adaptation, and training-based Conditional Critic Guidance (CCG) for amortized inference. Empirical evaluations demonstrate that our guided streaming approach significantly outperforms conventional chunk-based policies in reactivity and provides superior, physically valid guidance for dynamic, unstructured environments.

We introduce BEACON--Best-Effort Adaptation for Cross-Domain Co-Training--a theory-driven framework for training generative robot policies with abundant source demonstrations and limited target demonstrations. BEACON casts cross-domain co-training as a discrepancy-aware importance-reweighting problem, jointly learning a diffusion-based visuomotor policy and per-sample source weights that minimize an objective informed by target-domain generalization guarantees. To make best-effort adaptation practical for high-dimensional sequence policies, we develop scalable instance-level discrepancy estimators, stochastic alternating updates for policy and weights, and a multi-source extension that balances heterogeneous source domains. Across sim-to-sim, sim-to-real, and multi-source manipulation settings, BEACON improves robustness and data efficiency over target-only, fixed-ratio co-training, and feature-alignment baselines. Importantly, even without an explicit alignment objective, BEACON achieves feature alignment as an implicit result of discrepancy-aware cross-domain co-training.

While Vision-Language-Action (VLA) models offer broad general capabilities, deploying them on specific hardware requires real-world adaptation to bridge the embodiment gap. Since robot demonstrations are costly, this adaptation must often occur under a strict data budget. In this work, we identify a critical diversity trap: the standard heuristic of "maximizing coverage" by collecting diverse, single-shot demonstrations can be self-defeating due to non-vanishing estimation noise. We formalize this phenomenon as a Coverage--Density Trade-off. By decomposing the policy error into estimation (density) and extrapolation (coverage) terms, we characterize an interior optimal allocation of unique conditions for a fixed budget. Guided by this analysis, we propose Anchor-Centric Adaptation (ACA), a two-stage framework that first stabilizes a policy skeleton through repeated demonstrations at core anchors, then selectively expands coverage to high-risk boundaries via teacher-forced error mining and constrained residual updates. Real-robot experiments validate our trade-off framework and demonstrate that ACA significantly improves task reliability and success rates over standard diverse sampling strategies under the same budget.