Goal-conditioned world models transcribe instructions instead of perceiving spatial relations when the instruction names the scored quantity, and removing the goal from the dynamics fixes it.
Justin Johnson, Bharath Hariharan, Laurens van der Maaten, Li Fei-Fei, C
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Grounding Spatial Relations in a Compact World Model: Instruction Leakage and a Goal-Free Dynamics Fix
Goal-conditioned world models transcribe instructions instead of perceiving spatial relations when the instruction names the scored quantity, and removing the goal from the dynamics fixes it.