ProSR adds a Counterfactual Invariance Penalty and a Tail Drift Penalty to shape VLM reasoning trajectories for better visual dependence and stability on spatial tasks.
Entrocut: Entropy-guided adaptive truncation for efficient chain-of-thought reasoning in small-scale large reasoning models.arXiv preprint arXiv:2601.22617, 2026
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ProSR: Process-Shaped Spatial Reasoning for Reliable Chain-of-Thought in VLMs
ProSR adds a Counterfactual Invariance Penalty and a Tail Drift Penalty to shape VLM reasoning trajectories for better visual dependence and stability on spatial tasks.