ml-ai-agents-py/trm

tiny recursive models

• less is more: recursive reasoning with tiny networks, by a. jolicoeur-martineau (2025)
• • much simpler recursive reasoning approach that achieves significantly higher generalization than HRM on a net with only 2 layers.
  • with 7M parameters: 45% test-accuracy on ARC-AGI1, 8% on ARC-AGI-2 (higher than most LLMs with > 0.01% of the parameters).
  • "currently, recursive reasoning models such as HRM and TRM are supervised learning methods rather than generative models. this means that given an input question, they can only provide a single deterministic answer. in many settings, multiple answers exist for a question. thus, it would be interesting to extend TRM to generative tasks."