MiraBench: Evaluating Action-Conditioned Reliability in Robotic World Models

• URL: http://arxiv.org/abs/2605.29360v1
• Date: Thu, 28 May 2026 04:58:15 GMT
• Title: MiraBench: Evaluating Action-Conditioned Reliability in Robotic World Models
• Abstract summary: We introduce textscMiraBench, a hierarchical benchmark that defines emphaction-conditioned reliability as a core evaluation target for robotic world models.<n>To support this evaluation, we curate a human-annotated corpus with over 16,000 judgments across tasks, failure categories, and leading world models.<n>Across this broad model landscape, MiraBench reveals three central findings: visual fidelity is a poor proxy for action fidelity; increasing model scale does not reliably improve action following; and optimism bias is pervasive across current systems.
• Score: 25.87580992111249
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Action-conditioned world models are increasingly used as scalable simulators for robot learning, yet current evaluations provide limited evidence that their predictions are reliable under the actions they condition on. Existing benchmarks largely emphasize visual fidelity, leaving unclear whether predicted futures are physically plausible, faithful to commanded actions, and calibrated to failure when actions should not succeed. We introduce \textsc{MiraBench}, a hierarchical benchmark that defines \emph{action-conditioned reliability} as a core evaluation target for robotic world models. MiraBench decomposes this target into three progressively demanding levels: \emph{Physics Adherence}, which evaluates reference-free physical consistency; \emph{Action-Following Fidelity}, which measures whether predictions respect task-relevant action inputs; and \emph{Optimism Bias Detection}, which probes the tendency to predict successful outcomes under failure-inducing actions. To support this evaluation, we curate a human-annotated corpus with over 16,000 judgments across tasks, failure categories, and leading world models. We evaluate 12 representative model configurations spanning vector-conditioned robotic world models, text-conditioned generative world models, open-weight systems, closed-source systems, and multiple model scales. Across this broad model landscape, MiraBench reveals three central findings: visual fidelity is a poor proxy for action fidelity; increasing model scale does not reliably improve action following; and optimism bias is pervasive across current systems. By shifting evaluation from appearance to action-conditioned reliability, MiraBench provides a diagnostic foundation for assessing and improving robotic world models as faithful simulators.

Related papers

• Self-Evolving World Models for LLM Agent Planning [72.99782619992361]
  We introduce WorldEvolver, a self-evolving world model framework that revises its deployment-time context while keeping the downstream agent and all model parameters frozen.<n>WorldEvolver achieves the highest prediction accuracy across three backbones and leads other world model baselines on downstream agent success rate.
  arXiv  Detail & Related papers  (2026-06-29T17:58:43Z)
• Feedback World Model Enables Precise Guidance of Diffusion Policy [36.965417653906535]
  We propose feedback world model, a new paradigm that closes the loop between prediction and observation at inference time.<n>We show that our method substantially improves both prediction accuracy and policy performance under distribution shift.<n>In particular, it reduces world model prediction error by up to 76.4% and improves out-of-distribution (OOD) success rate by 30%.
  arXiv  Detail & Related papers  (2026-05-15T07:52:13Z)
• World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry [82.93104394404781]
  General-purpose world models promise scalable policy evaluation, optimization, and planning.<n>We propose World Action Verifier (WAV), a framework that enables world models to identify their own prediction errors and self-improve.
  arXiv  Detail & Related papers  (2026-04-02T12:48:36Z)
• ManipArena: Comprehensive Real-world Evaluation of Reasoning-Oriented Generalist Robot Manipulation [66.02142169323521]
  Vision-Language-Action models and world models have recently emerged as promising paradigms for general-purpose robotic intelligence.<n>Existing benchmarks are largely simulator-centric, which provide controllability but fail to capture the reality gap caused by perception noise.<n>We introduce ManipArena, a standardized evaluation framework designed to bridge simulation and real-world execution.
  arXiv  Detail & Related papers  (2026-03-30T15:06:41Z)
• WorldArena: A Unified Benchmark for Evaluating Perception and Functional Utility of Embodied World Models [114.95269118652163]
  We introduce WorldArena, a unified benchmark designed to evaluate embodied world models across both perceptual and functional dimensions.<n>WorldArena assesses models through three dimensions: video perception quality, measured with 16 metrics across six sub-dimensions; embodied task functionality, which evaluates world models as data engines, policy evaluators, and action planners integrating with subjective human evaluation.<n>Through extensive experiments on 14 representative models, we reveal a significant perception-functionality gap, showing that high visual quality does not necessarily translate into strong embodied task capability.
  arXiv  Detail & Related papers  (2026-02-09T18:09:20Z)
• WorldLens: Full-Spectrum Evaluations of Driving World Models in Real World [100.68103378427567]
  Generative world models are reshaping embodied AI, enabling agents to synthesize realistic 4D driving environments that look convincing but often fail physically or behaviorally.<n>We introduce WorldLens, a full-spectrum benchmark evaluating how well a model builds, understands, and behaves within its generated world.<n>We further construct WorldLens-26K, a large-scale dataset of human-annotated videos with numerical scores and textual rationales, and develop WorldLens-Agent.
  arXiv  Detail & Related papers  (2025-12-11T18:59:58Z)
• WorldSimBench: Towards Video Generation Models as World Simulators [79.69709361730865]
  We classify the functionalities of predictive models into a hierarchy and take the first step in evaluating World Simulators by proposing a dual evaluation framework called WorldSimBench. WorldSimBench includes Explicit Perceptual Evaluation and Implicit Manipulative Evaluation, encompassing human preference assessments from the visual perspective and action-level evaluations in embodied tasks. Our comprehensive evaluation offers key insights that can drive further innovation in video generation models, positioning World Simulators as a pivotal advancement toward embodied artificial intelligence.
  arXiv  Detail & Related papers  (2024-10-23T17:56:11Z)
• Zero-shot Safety Prediction for Autonomous Robots with Foundation World Models [0.12499537119440243]
  A world model creates a surrogate world to train a controller and predict safety violations by learning the internal dynamic model of systems. We propose foundation world models that embed observations into meaningful and causally latent representations. This enables the surrogate dynamics to directly predict causal future states by leveraging a training-free large language model.
  arXiv  Detail & Related papers  (2024-03-30T20:03:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.