After length correction, reasoning-trained language models exhibit distinct hidden-state trajectory geometries on harder problems compared to instruction-tuned baselines, with the strongest effect in code domains.
LLMs encode how difficult problems are.arXiv preprint arXiv:2510.18147
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Reasoning Models Don't Just Think Longer, They Move Differently
After length correction, reasoning-trained language models exhibit distinct hidden-state trajectory geometries on harder problems compared to instruction-tuned baselines, with the strongest effect in code domains.