Equilibrium Reasoners learn task-conditioned attractors in latent dynamics to support scalable iterative reasoning, raising Sudoku-Extreme accuracy from 2.6% to over 99% via up to 40,000 equivalent layers.
International Conference on Learning Representations , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning
Equilibrium Reasoners learn task-conditioned attractors in latent dynamics to support scalable iterative reasoning, raising Sudoku-Extreme accuracy from 2.6% to over 99% via up to 40,000 equivalent layers.