AHA · LOOPED TRANSFORMERSubmit a paper ↗
← Back to the collection
Adaptive depth ·

ERM

Energy-guided Recursive Model

Yifei Zhao, Ying Tang

Energy-guided Recursive Model (ERM) augments recurrent reasoning with an explicit energy-based mechanism for selecting among candidate trajectories. It constructs task-structured Hopfield energies over valid local or global patterns, using them to rank recurrent outputs and to support energy-guided sampling such as parallel tempering.

Inside the method

EqR-style recurrent reasoner × D → K candidate trajectories → Hopfield-energy selector → selected answer

Simplified conceptual schematic. Consult the paper for the complete architecture.

Recurrence family
Adaptive depth
Depth control
D=64 recurrent steps in the main experiments; test-time computation scales through recurrent depth and candidate breadth
KV / state strategy
See paper

Reading note

ERM primarily changes the inference-time selection mechanism rather than replacing the recurrent reasoning backbone. The main experiments use EqR-style recurrent rollouts with D=64 and K=128 candidates. The energy is task-structured: Sudoku uses row, column, and box permutation memories, PPBench uses puzzle-specific local memories and global distance potentials, and the reported Maze global-memory setting uses an input-conditioned shortest-path memory.

Sources checked 2026-09-16. This catalog does not imply independent reproduction.

Cite this work

@misc{arxiv2607101282026,
  title = {Energy-guided Recursive Model},
  author = {Yifei Zhao and Ying Tang},
  year = {2026},
  eprint = {2607.10128},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2607.10128}
}

Suggest a correction ↗