Physically Viable World Models: A Case for Query-Conditioned Embodied AI
Quick Take
This paper argues for the necessity of physically viable world models in embodied AI, emphasizing that models must accurately represent physical structures to respond to intervention queries. It critiques existing observation-predictive models for their structural failures and proposes a modular approach to model design that enhances interpretability and verifiability, ensuring safer and more feasible action recommendations.
Key Points
- Existing models often produce physically incorrect predictions despite visual plausibility.
- Controlled benchmarks reveal failures in recommending feasible actions and predicting outcomes.
- Proposed models should include modular components for environment representation and action specification.
- The right abstraction is the simplest model preserving relevant distinctions for queries.
- Dynamic assembly of models can improve planning, control, and verification in AI systems.
Article Content
From source RSS / original summaryarXiv:2605. 30542v1 Announce Type: new Abstract: World models for embodied AI must be physically viable: constructed to answer intervention queries by representing the physical structure governing action outcomes, rather than merely predicting future observations. Existing observation-predictive world models can produce visually plausible but physically wrong rollouts. This failure is structural; distinct physical systems can look identical yet diverge under intervention.
We expose this problem with controlled benchmarks that fix the visible scene while varying latent physics. We show that such models may recommend infeasible actions, mispredict interaction outcomes, or certify unsafe behavior. We argue that embodied AI requires world models that identify the simplest physical abstraction sufficient to answer an intervention query.
Such a model comprises modular components, including environment representation, latent state and parameter estimation, action specification, interventional dynamics, and query-level response. An autonomous orchestrator should identify the relevant abstraction and compose compatible learned and structured components per query.
When closed-form physics is unavailable, uncertain, or costly, the transition model may be analytic, simulated, learned, or hybrid, but it must preserve the structure that determines interventional outcomes. This decomposition makes the model interpretable, its components verifiable, and its outputs auditable against the query.
It also provides a design principle for new world models and a feasibility test for existing ones: the right abstraction is not the most detailed model of the world, but the simplest model that preserves the distinctions relevant to the query. We demonstrate this approach on queries that existing systems fail to answer correctly, and outline how an orchestrator can dynamically assemble and adapt physically viable models for planning, control, and verification.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →The Importance of Out-of-Band Metadata for Safe Autonomous Agents: The Redpanda Agentic Data Plane
The Redpanda Agentic Data Plane (ADP) introduces out-of-band metadata channels to enhance the safety of autonomous AI agents, ensuring secure data access and tamper-proof audit trails. This architecture mitigates risks associated with unpredictable AI behavior by enforcing governance throughout the agent lifecycle, demonstrated in a multi-agent trading system with strict data scoping and approval thresholds.
