import torch

# Load checkpoint
ckpt = torch.load("trm_fullbp/step_9764", map_location="cpu")

# Load into TRM model (requires TinyRecursiveModels codebase)
# model = TinyRecursiveReasoningModel_ACTV1(config)
# model.load_state_dict(ckpt, strict=False)

Paper

Disentangling Gradient Quality from Architecture in Recursive Reasoning Models | Code

Abstract: We disentangle gradient quality from architecture design in recursive reasoning models by running TRM's flat architecture with HRM's 1-step gradient approximation. Our results show gradient quality, not architecture, is the dominant factor explaining the performance gap.

Citation

bibtex

@article{jj2026disentangling,
  title={Disentangling Gradient Quality from Architecture in Recursive Reasoning Models},
  author={Jatin (JJ)},
  doi={10.5281/zenodo.20712090},
  year={2026}
}

heyiamjj/disentangling-gradients-recursive-reasoning

Disentangling Gradient Quality from Architecture in Recursive Reasoning Models

Models

Architecture

Training

Usage

Paper

Citation

heyiamjj/
disentangling-gradients-recursive-reasoning