# Dual-Frontier: When Can an Agent Trust Its World Model?

arXiv: https://arxiv.org/abs/2609.26293 · HTML: https://mlsanity.com/paper/2609.26293/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in World models for language agents, one of 11 directions in World models.
- Builds on 13 of its 45 references that are in the map.
- First posted on arXiv 22 Sep 2026.

## Placed in

- World models / [World models for language agents](https://mlsanity.com/wm/world-models-for-language-agents/)

## Builds on (most cited first, 13 of 13)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [GAIA-2: A Controllable Multi-View Generative World Model for Autonomous Driving](https://mlsanity.com/paper/2503.20523/) (2025)
- [Agent World Model: Infinity Synthetic Environments for Agentic Reinforcement Learning](https://mlsanity.com/paper/2602.10090/) (2026)
- [Qwen-AgentWorld: Language World Models for General Agents](https://mlsanity.com/paper/2606.24597/) (2026)
- [Policy-Aware Model Learning for Policy Gradient Methods](https://mlsanity.com/paper/2003.00030/) (2020)
- [Policy and World Modeling Co-Training for Language Agents](https://mlsanity.com/paper/2606.02388/) (2026)
- [Self-Evolving World Models for LLM Agent Planning](https://mlsanity.com/paper/2606.30639/) (2026)
- [CoMAP: Co-Evolving World Models and Agent Policies for LLM Agents](https://mlsanity.com/paper/2606.02372/) (2026)
- [General Agents Contain World Models, even under Partial Observability and Stochasticity](https://mlsanity.com/paper/2602.03146/) (2026)
- [Tools as Continuous Flow for Evolving Agentic Reasoning](https://mlsanity.com/paper/2605.07339/) (2026)

## Built on by (most cited first, 1 of 1)

- [When Low Prediction Error Misleads Planning: Diagnosing Representation, Dynamics, and Decision Failures in Latent World Models](https://mlsanity.com/paper/2610.05550/) (2026)

## Abstract (arXiv)

Learned world models are becoming essential to general-purpose agents: by predicting action consequences, they support planning and decision-making while reducing reliance on costly trial and error. This reliance creates a fundamental ambiguity: when a world-model-guided decision fails, the trajectory alone may not reveal whether the agent's decision rule or the world model caused the loss. We formalize this failure-attribution problem as a counterfactual decomposition of return loss and prove that its components are not identifiable from passive interaction, even for finite-horizon planners. This obstruction motivates Dual-Frontier, a learning principle that admits a world-model-guided decision only when its predicted advantage exceeds a certified bound on decision-relevant world-model error; otherwise, evidence is allocated to world-model verification. Action-conditioned value bounds and a closed-loop extension guarantee non-decreasing return for admitted decisions. Calibrated gates and simultaneous confidence sequences support adaptive evidence reuse, with sufficient and necessary verification bounds. Controlled learned-model experiments validate the predicted failure modes and certification behavior, while cross-backbone tool-use benchmarks instantiate the same verify-then-promote rule in realistic agent world-model pipelines, consistently improving decision quality and reliability.