cs.AINov 12, 2026

Classical Energies Learn to Reason 1,000x Faster than Looped Models

Authors: Benjamin Hoover, Hugh Blayney, James C. R. Whittington

Organizations: Prefrontal, London, United Kingdom

Abstract

Looped models have emerged as a prominent paradigm for reasoning, with successes on simple reasoning puzzles such as Sudoku-Extreme and Maze-Unique. However, these models can be unstable, slow to train, and use many parameters and FLOPs to solve simple puzzles. Descent on an energy not only mitigates these issues but also offers advantages: built-in early stopping, compute that adapts to problem difficulty, fast training without differentiating through a solver, and interpretable weights that learn the constraints of the puzzle. Here we demonstrate this with a learned version of Hopfield & Tank's energy from the 80s, which we term CRAM, solves 99.996% of Sudoku-Extreme and 100% of Maze-Unique test puzzles after 42 ms and 132 ms of training on one GPU, at least 1,000x faster than looped models on these tasks. CRAM also matches baseline looped models at their highest accuracy on Sudoku-Extreme using at least 282x fewer inference FLOPs. Our work suggests two things. First, that energies are an extremely promising direction for training reasoning from data. Second, a caution against using trivially solvable benchmarks like Sudoku and Maze to test the scalability of looped reasoning techniques.

Figures & tables

Appendix figures & tables7 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Energy-Based Reasoners: Your Recurrent Reasoner Should Be an Energy-Based Model

    Nov 6, 2026William Xingxu Chen, Alexi Gladstone, Zeyu Michael Li +2Recursive Reasoning ModelsEfficient Language Model Reasoning

  2. Flow Reasoning Models: Turning Flows Into Efficient Recurrent Reasoners

    Jun 28, 2026Alec Helbling, Andrey Bryutkin, Mauro Martino +3Self-ConditioningStructured Reasoning

  3. Resample the Answer: Where to Spend Test-Time Compute in Recursive Reasoning

    Oct 12, 2026Maxime IllouzRecursive Reasoning ModelsInference-Time Search