Reliable performance evaluation is a central bottleneck for deploying robot-learning policies in real-world conditions. Real-world testing is faithful but costly and difficult to scale, whereas simulation-based testing scales easily but is inevitably biased by the sim-to-real gap. Existing simulation-augmented methods combine limited real-world rollouts with abundant simulation proxies, but focus on performance averaged over initial conditions and deployment settings. Such population-level averages obscure scenario-specific variation and provide limited guidance about when and where a policy can be safely deployed. We propose SCAPE, a scenario-conditioned simulation-augmented policy evaluation framework that predicts scenario-conditioned real-world policy performance using limited paired sim-and-real samples and large-scale simulation rollouts. SCAPE corrects sim-to-real bias in simulation labels before training the prediction model and calibrates prediction uncertainty through conformal prediction. We validate SCAPE on autonomous driving and quadruped velocity tracking. In sim-to-sim studies, SCAPE reduces scenario-level prediction error by 4.9%/34.7% (driving) and 14.5%/27.7% (quadruped) relative to scene-conditioned neural and aggregate statistical baselines on average. We further evaluate a velocity-tracking policy deployed on a physical Unitree Go2. SCAPE also improves testing sample efficiency, produces narrower calibrated prediction intervals, generalizes better to out-of-distribution scenarios, and enables fine-grained deployment strategies.
Reliable robot learning requires a world simulator that can predict action consequences before execution on physical hardware, including risky and failure-prone outcomes. Existing physics simulators require substantial asset construction and calibration and still face a sim-to-real gap, while video generators often lack precise control over their responses to fine-grained robot actions. In this paper, we present the Boundless World Model (BWM), an open-source, low-cost, high-fidelity world simulator for robot manipulation. BWM is an action-conditioned world model that combines initial-environment guidance, dynamic visual history, and temporally aligned robot-action conditioning for stateful autoregressive prediction of future observations. We construct action-aligned training clips through trajectory replay, overlapping clip sampling, and initial-observation enhancement. BWM serves as a data engine that augments imitation-learning data with action-aligned rollouts, and as a policy evaluator for closed-loop assessment, risk anticipation, and policy ranking. Experiments on the WorldArena benchmark and physical robots demonstrate improved simulator fidelity and functional utility across the data-engine and policy-evaluator settings. BWM ranks first overall in the WorldArena Challenge across Track 1 and its two Track 2 applications. We release the BWM open-source ecosystem, including model checkpoints, training and inference code, and interfaces for data generation and policy evaluation.
Action-conditioned world models are a key component of embodied AI, serving as scalable policy evaluators that reduce reliance on expensive real-world rollouts. To accurately capture diverse action-induced dynamics, such models should satisfy three key objectives-Physical Plausibility (P), Action Adherence (A), and Visual Fidelity (V), collectively referred to as PAV-while remaining robust to both in-distribution (ID) expert demonstrations and out-of-distribution (OOD) actions. However, existing methods primarily rely on ID action-video pairs and pixel-level reconstruction losses, which do not explicitly optimize PAV objectives and generalize poorly beyond expert data. To address this, we propose PAVXploreRL, a reinforcement learning framework built on a pretrained latent world model that explicitly optimizes PAV objectives through reward-driven training. To improve action generalization, our method jointly leverages ID trajectories and noise-driven OOD action exploration, without paired video supervision. Experiments show that PAVXploreRL consistently outperforms pretrained baselines, achieving a 5.6% average gain across benchmarks and producing higher-quality PAV properties. As a policy evaluator, it also yields more reliable performance estimates and reduces the overestimation bias of prior expert-only world models such as Ctrl-World. Code: https://github.com/Social-AI-Studio/PAVXploreRL
Real-world evaluation is the gold standard for robot policies because it tests them against the physical conditions and deployment challenges they are ultimately designed to handle. However, real-world evaluation is also the bottleneck for iterating on robot policies: it is costly, difficult to reproduce, and often too sparse to reliably compare nearby model variants. A straightforward proxy for performance is validation loss on expert demonstrations, but this proxy is often poorly correlated with real-world performance. In this paper, we introduce Critical Interval MSE (CI-MSE), an intuitively simple yet effective offline validation metric. CI-MSE restricts error computation to task-critical segments and pairs it with simple action-alignment procedures that better match rollout-time behavior. Across simulation and real-world experiments, CI-MSE yields a stronger correlation between validation error and rollout performance than raw MSE. Across a wide range of policy checkpoints, CI-MSE achieves a Spearman's rank correlation of $-0.87$, much closer to the ideal value of $-1$ than raw MSE's $-0.61$, demonstrating a significant improvement. We show through sensitivity analysis that our metric is robust to a wide range of hyperparameters. We further study the effectiveness of CI-MSE under evaluation distribution shifts and suggest design boundaries when using this metric. In summary, this paper provides a simple and reliable offline validation tool for accelerating policy iteration. Project webpage: https://ci-mse.github.io/
Evaluating generalist robot manipulation policies in the real world is expensive, slow, and difficult to scale. Action-conditioned video world models offer a scalable alternative by simulating policy rollouts. Autoregressive rollouts accumulate compounding errors, observations across multiple camera views must remain mutually consistent, and the evaluator must generalize to policies whose behaviors lie outside the training distribution. We address these challenges with SC3-Eval, a self-consistent video generation recipe that adapts a pre-trained video foundation model into an accurate policy evaluator by enforcing three complementary forms of consistency. First, forward-inverse dynamics consistency jointly trains the model to predict frames from actions and to recover actions from frames, anchoring generated rollouts to a physically plausible action manifold and counteracting the drift a forward-only model cannot penalize. Second, cross-view consistency trains the model to inpaint each camera view from the other, keeping the multi-camera observation coherent over long rollouts without any explicit memory mechanism. Third, test-time consistency reuses the inverse dynamics mode at inference as a per-action-chunk uncertainty signal that terminates rollouts whose generated frames drift away from the requested actions. We also demonstrate SC3-Eval rollouts reproduce the failure modes that policies exhibit in real-world rollouts, supporting fine-grained diagnostic comparison rather than aggregate ranking alone. Across seven real-world vision-language-action policies, SC3-Eval attains a closed-loop Pearson correlation of $0.929$ and MMRV of $0.119$, outperforming three strong prior video-model-based baselines, and generalizes to new tasks.
We study whether demonstration-curation metrics that detect defective training episodes also improve the downstream behavior-cloning policy that trains on the curated data. On a contact-rich LIBERO pick-and-place benchmark with a controlled structural defect (early gripper release during the carry phase), we find that the two quantities are sharply decoupled. The metric with the highest defect-detection AUROC (0.804) produces the worst curated policy (13.3% task success), while a metric with a substantially lower AUROC (0.638) produces a policy that nearly matches the oracle trained on ground-truth clean data (90.0% vs. 93.3%). We further show that five of the seven metrics we evaluate exploit episode length as a trivial proxy for the defect label, a confound that inflates reported AUROCs to near-perfect values and disappears once episode length is controlled. Across all conditions, the contaminated baseline succeeds on only 3.3% of rollouts, and the two best curation methods close this to within 3 percentage points of the 93.3% oracle ceiling. Our results argue that curation methods should be evaluated by the policy they produce, not the defects they flag, and that any curation benchmark must control for episode length before reporting detection accuracy. We release the testbed, all metric implementations, and the evaluation pipeline.
Skill libraries in deployed robotic systems are continually updated through fine-tuning, fresh demonstrations, or domain adaptation, yet existing typed-composition methods (BLADE, SymSkill, Generative Skill Chaining) treat the library as frozen at test time and do not analyze how composition outcomes change when a skill is replaced. We introduce a paired-sampling cross-version swap protocol on robosuite manipulation tasks to characterize this dimension of compositional skill learning. On a dual-arm peg-in-hole task we discover a dominant-skill effect: one ECM achieves 86.7% atomic success rate while every other ECM is at or below 26.7%, and whether this dominant ECM enters a composition shifts the success rate by up to +50pp. We characterize the boundary on a simpler pick task where all atomic policies saturate at 100% and the effect is undefined. Across three tasks we further find that off-policy behavioral distance metrics fail to identify the dominant ECM, ruling out the natural cheap predictor. We propose an atomic-quality probe and a Hybrid Selector combining per-skill probes (zero per-decision cost) with selective composition revalidation (full cost), and characterize its Pareto frontier on 144 skill-update decisions. On T6 the atomic-only probe sits 23pp below full revalidation (64.6% vs 87.5% oracle match) at zero per-decision cost; a Hybrid Selector with m=10 closes most of that gap to ~12pp at 46% of full-revalidation cost. On the cross-task average over 144 events, atomic-only is within 3pp of full revalidation under a mixed-oracle caveat. The atomic-quality probe is, to our knowledge, the first principled, deployment-ready primitive for skill-update governance in compositional robot policies.