Robotic Policy Evaluation

Video world models (WMs) have shown promise for policy evaluation and improvement by imagining realistic future observations conditioned on ego-robot actions. While WMs can model distributions over futures, policy evaluation and improvement typically rely on nominal imaginations, which can miss high-impact outcomes of robot actions unless prohibitively many samples are drawn. To enable robust policy evaluation and improvement over WM imaginations, we propose StressDream, which steers imaginations toward high-impact yet plausible outcomes specified at inference time by optimizing the initial noise of diffusion-based WMs. However, optimizing high-dimensional noise is challenging: the optimization must reason about nuanced, scene-dependent target events in generated videos while avoiding out-of-distribution (OOD) noise that yields implausible imaginations. We address this with two complementary objectives: a semantic objective with a Vision-Language Model that provides informative gradients by reasoning about the generated video, and a plausibility objective that prevents the optimized noise from drifting OOD. With state-of-the-art video world models for autonomous driving and robotic manipulation, we show that StressDream effectively steers imaginations toward high-impact yet plausible outcomes specified by text at inference time, such as task failures, enabling robust policy evaluation and improvement by identifying actions whose plausible futures include undesirable outcomes. Video results are available at https://stressdream.github.io/.

Fine-tuning Vision-Language-Action (VLA) models for mobile manipulators with heterogeneous joint spaces can produce a counterintuitive result: the checkpoint with the lowest aggregate MSE is not the one that performs best on the real robot. We argue this is a predictable consequence of collapsing heterogeneous joint groups (arm, gripper, head, wheeled base) into a single metric, where easy-to-predict joints can mask joints that still fail. We fine-tune SmolVLA (450M, action-expert only) on the 11-DoF Toyota HSR and compare it against $π_{0.5}$ (3.3B), a stronger pretrained baseline. Per-group analysis exposes two patterns: in SmolVLA, the mobile base converges slowest and limits overall performance. In expert-only fine-tuning of $π_{0.5}$ (training only the action head, backbone frozen), total MSE drops below the baseline but arm accuracy degrades. On 60 real-robot trials (20 per model), $π_{0.5}$ 80k (4.0/4) significantly outperforms both fine-tuned variants (expert-only 3k: 3.75/4; HSR-SmolVLA: 3.5/4; Mann-Whitney $p \leq 0.010$), despite expert-only 3k having the lowest total MSE. This separation is most consistent with the offline arm-group error, not total MSE or base-group error. We conclude that per-group error is a more reliable signal than total MSE for checkpoint selection on robots with heterogeneous action spaces. Code: https://github.com/paumontagut/per-group-mse-vla

Real-world evaluation of vision-language-action (VLA) policies still rests on binary success rate at a fixed timeout with $N \le 25$ rollouts per condition, almost always without confidence intervals or paired statistical comparison; these cohort sizes struggle to resolve close comparisons reliably. We introduce PhAIL (Physical AI Leaderboard, https://phail.ai), an open real-robot benchmark on a Franka FR3 (dataset, per-rollout artifacts, and end-to-end reference implementation) of a distributional evaluation methodology: the time-to-success cumulative distribution function (CDF) as the evaluation primitive, with two separated jobs. The first is scoring via Human-Relative Throughput (HRT), a dimensionless scalar with bootstrap confidence intervals, anchored to same-fixture human teleoperation. The second is a significance test (Kolmogorov-Smirnov, computed per-object and macro-averaged across objects). On four publicly-available VLAs, the macro-averaged KS test resolves two close comparisons (GR00T vs. ACT, OpenPI vs. ACT) at $N \le 30$ rollouts per (model, object) cell where binary-threshold metrics do not; the closest pair (OpenPI vs. GR00T) remains unresolved within our budget. The best evaluated VLA is $\sim 7\times$ slower per operation (RMST ratio) than the human reference.

Robot learning research is fragmented across policy families, benchmark suites, and real robots; each implementation is entangled with the others in a complex combination matrix, making it an engineering nightmare to port any single element. General-purpose coding agents may occasionally bridge specific setups, but cannot close this gap at scale because they lack the procedural priors and validation practices that characterize robotics research workflows. We propose NAUTILUS, an open-source harness that turns a single user prompt -- for example, "Evaluate policy A with benchmark B" -- into ready-to-use reproduction, evaluation, fine-tuning, and deployment workflows. NAUTILUS provides: plug-and-play agent skill sets with distilled priors from robotics research; typed contracts among policies, simulators/benchmarks, and real-world robots; unified interfaces and execution environments; and a trustworthy agentic coding workflow with explicit, automated validation, and testing at each milestone. NAUTILUS can not only automatically generate the required adapters and containers for existing implementations, but also wrap and onboard new or user-provided policies, simulators/benchmarks, and robots, all connected via a uniform interface. This expands cross-validation coverage without hand-written glue code. Like a nautilus shell that grows by adding chambers, NAUTILUS scales by extending its execution in chambered units, making it a research harness for scalability rather than a hand-curated framework, and aiming to reduce the engineering burden of cross-family reproduction and evaluation in the ever-growing robot learning ecosystem.

Policy evaluation is a fundamental component of the development and deployment pipeline for robotic policies. In modern manipulation systems, this problem is particularly challenging: rewards are often sparse, task progression of evaluation rollouts are often non-monotonic as the policies exhibit recovery behaviors, and evaluation rollouts are necessarily of finite length. This finite length introduces truncation bias, breaking the infinite-horizon assumptions underlying standard methods relying on Bellman equations/principle of optimality. In this work, we propose a framework for offline policy evaluation from sparse rewards based on a liveness-based Bellman operator. Our formulation interprets policy evaluation as a task-completion problem and yields a conservative fixed-point value function that is robust to finite-horizon truncation. We analyze the theoretical properties of the proposed operator, including contraction guarantees, and show how it encodes task progression while mitigating truncation bias. We evaluate our method on two simulated manipulation tasks using both a Vision-Language-Action model and a diffusion policy, and a cloth folding task using human demonstrations. Empirical results demonstrate that our approach more accurately reflects task progress and substantially reduces truncation bias, outperforming classical baselines such as TD(0) and Monte Carlo policy evaluation.

Evaluating robotics policies across thousands of environments and thousands of tasks is infeasible with existing approaches. This motivates the need for a new methodology for scalable robotics policy evaluation. In this paper, we propose dWorldEval, which uses a discrete diffusion world model as a scalable evaluation proxy for robotics policies. Specifically, dWorldEval maps all modalities - including vision, language, and robotic actions - into a unified token space, modeling them via a single transformer-based denoising network. In this paper, we propose dWorldEval, using a discrete diffusion world model as a scalable evaluation proxy for robotics policy. Specifically, it maps all modalities, including vision, language, and robotics action into a unified token space, then denoises them with a single transformer network. Building on this architecture, we employ a sparse keyframe memory to maintain spatiotemporal consistency. We also introduce a progress token that indicates the degree of task completion. At inference, the model jointly predicts future observations and progress token, allowing automatically determine success when the progress reaches 1. Extensive experiments demonstrate that dWorldEval significantly outperforms previous approaches, i.e., WorldEval, Ctrl-World, and WorldGym, on LIBERO, RoboTwin, and multiple real-robot tasks. It paves the way for a new architectural paradigm in building world simulators for robotics evaluation at scale.

Robotic manipulation policies often degrade over extended horizons, yet existing benchmarks provide limited insight into why such failures occur. Most prior benchmarks are either simulation-based or report aggregate success, making it difficult to disentangle the distinct sources of temporal difficulty in real-world execution. We introduce LongBench, a real-world benchmark for evaluating long-horizon manipulation. LongBench consists of over 1,000 real-world episodes, covering two complementary regimes: Context-Independent (fully observable) and Context-Dependent (ambiguity-driven). By organizing tasks into capability- and ambiguity-specific subsets, LongBench enables mechanism-aware evaluation of execution robustness, temporal consistency, and context-dependent reasoning. Evaluating six state-of-the-art policies reveals that long-horizon performance is not governed by a single factor. We observe that performance in fully observable settings is more strongly associated with execution robustness, while contextual difficulty varies across tasks and is not consistently improved by memory-based methods. We hope that LongBench serves as a useful benchmark for studying long-horizon manipulation and for developing policies with stronger robustness across both execution and contextual challenges.

Although Vision--Action (VA) and Vision--Language--Action (VLA) policies have advanced robotic manipulation, their evaluation remains dominated by binary success rates, which obscure process-level differences among executions that complete the same task. We introduce Eval-Actions, a diagnostic evaluation methodology and real-robot benchmark for fine-grained execution-quality assessment of learned manipulation policies. Eval-Actions combines criteria-based Expert Grading (EG), Rank-Guided (RG) labels that align measurable motion indicators with expert rankings, and Chain-of-Thought-style (CoT) annotations that explain observable quality differences. The benchmark contains 13K+ teleoperated and policy-generated real-robot episodes covering 150+ tasks and approximately 52 hours of recordings with RGB-D videos, robot-state trajectories, task descriptions, and success/failure labels. Its densely annotated subset provides EG/RG/CoT supervision for training and evaluation. We further provide AutoEval, a reference multimodal evaluator that predicts quality scores, task outcomes, and diagnostic explanations from RGB temporal evidence and compact kinematic summaries. On the annotated Eval-Actions test split, AutoEval-S achieves Spearman rank correlations (SRCCs) of 0.81 and 0.84 under EG and RG, with success detection accuracies of 90.6% and 91.0%; AutoEval-P reaches 0.70 SRCC under CoT. Analyses of expert consistency, physical-metric baselines, modality ablations, structured generalization, and offline policy ranking show that Eval-Actions provides standardized, interpretable diagnostic signals complementary to success-rate evaluation.

This paper presents a quantitative comparison between centralised and distributed multi-agent reinforcement learning (MARL) architectures for controlling a soft robotic arm modelled as a Cosserat rod in simulation. Using PyElastica and the OpenAI Gym interface, we train both a global Proximal Policy Optimisation (PPO) controller and a Multi-Agent PPO (MAPPO) under identical budgets. Both approaches are based on the arm having $n$ number of controlled sections. The study systematically varies $n$ and evaluates the performance of the arm to reach a fixed target in three scenarios: default baseline condition, recovery from external disturbance, and adaptation to actuator failure. Quantitative metrics used for the evaluation are mean action magnitude, mean final distance, mean episode length, and success rate. The results show that there are no significant benefits of the distributed policy when the number of controlled sections $n\le4$. In very simple systems, when $n\le2$, the centralised policy outperforms the distributed one. When $n$ increases to $4< n\le 12$, the distributed policy shows a high sample efficiency. In these systems, distributed policy promotes a stronger success rate, resilience, and robustness under local observability and yields faster convergence given the same sample size. However, centralised policies achieve much higher time efficiency during training as it takes much less time to train the same size of samples. These findings highlight the trade-offs between centralised and distributed policy in reinforcement learning-based control for soft robotic systems and provide actionable design guidance for future sim-to-real transfer in soft rod-like manipulators.

We present EBench, a simulation benchmark that diagnoses generalist mobile manipulation policies beyond a single success-rate scalar. EBench comprises 26 diverse and challenging manipulation tasks annotated along 5 capability dimensions and 4 generalization dimensions. We evaluate state-of-the-art generalist manipulation models including $\pi_0$, $\pi_{0.5}$, XVLA, and InternVLA-A1, and reveal that the models exhibit strikingly different capability profiles: $\pi_{0.5}$ achieves the highest test success rate, the best train--test retention, and the strongest mobile manipulation performance; $\pi_0$ leads on dexterous fixed-base and high-precision tasks; XVLA and InternVLA-A1 exhibit complementary strengths across atomic skills and operating regimes. Beyond capability profiling, EBench analyzes the generalization ability from 4 representative perspectives, identifying the impact of different distribution shift factors. The results reveal strengths and weaknesses of models behind an overall score. We hope this benchmark offers a broad set of diagnostic signals to guide iteration on generalist manipulation models.