A depth-aware structured chain-of-thought trained with multi-objective process rewards improves VLM spatial reasoning benchmarks, though headline numbers are weakened by evaluation overlap and test-set tuning.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
SCOUT: Unlocking Enhanced Spatial Reasoning via Structured Chain-of-Thought and Multi-Objective Process Reward
A depth-aware structured chain-of-thought trained with multi-objective process rewards improves VLM spatial reasoning benchmarks, though headline numbers are weakened by evaluation overlap and test-set tuning.