Beyond Task Success: Stage-Wise Reliability of World Model Planning under Sensing Degradation
Classification
View four quadrants- Major category
- Evaluation metrics
- Quadrant
- Not applicable
- Architecture
- Not applicable
- Prediction paradigm
- Not applicable
- Source review status
- Not assigned
Category review. This is a stage-wise diagnostic and evaluation protocol for representation shift, prediction discrepancy, candidate ranking and executed outcomes under sensing corruption; it does not propose a new controller or reusable component. The cataloged contribution is a dataset/data-generation method or evaluation/simulation resource, not the architecture of an evaluated or external policy. Reading evidence
Contribution
This diagnostic study follows synthetic sensing disturbances through an existing world-model planner's representation, action-conditioned prediction, candidate ranking, and executed task outcome. Paired OGBench evaluations show that a disturbance's relative impact can change between stages. Temporal information position matters even when aggregate representation shifts are similar. These diagnostics locate sensitivity; they neither establish a general predictor of task failure nor demonstrate successful sensing mitigation.
Abstract
An abstract has not been added yet.
Affiliations
Seoul National University