mlsanityData to 7 Oct 2026

Less is More: Recursive Reasoning with Tiny Networks

arXiv 2510.04871 ↗First posted 6 Oct 2025cs.LG cs.AIcited by 38

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

Hierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on hard puzzle tasks such as Sudoku, Maze, and ARC-AGI while trained with small models (27M parameters) on small data (around 1000 examples). HRM holds great promise for solving hard problems with small networks, but it is not yet well understood and may be suboptimal. We propose Tiny Recursive Model (TRM), a much simpler recursive reasoning approach that achieves significantly higher generalization than HRM, while using a single tiny network with only 2 layers. With only 7M parameters, TRM obtains 45% test-accuracy on ARC-AGI-1 and 8% on ARC-AGI-2, higher than most LLMs (e.g., Deepseek R1, o3-mini, Gemini 2.5 Pro) with less than 0.01% of the parameters.

Key facts. Cited by 38 papers in the map (data through 7 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 15 of its 38 references that are in the map. First posted on arXiv 6 Oct 2025.

Builds on

15 of 38 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025Hierarchical Reasoning Model26 Jun
  2. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug

Other references · 13

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020Scaling Laws for Neural Language Modelscited by 6,411
  3. 2016Gaussian Error Linear Units (GELUs)cited by 3,431
  4. 2017Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layercited by 3,413
  5. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,609
  6. 2020GLU Variants Improve Transformercited by 1,724
  7. 2017Decoupled Weight Decay Regularizationcited by 1,722
  8. 2025ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systemscited by 171
  9. 2019On the Measure of Intelligencecited by 129
  10. 2023The ConceptARC Benchmark: Evaluating Understanding and Generalization in the ARC Domaincited by 97
  11. 2024Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrappingcited by 89
  12. 2025Grokking at the Edge of Numerical Stabilitycited by 41
1 more
  1. 2023TorchDEQ: A Library for Deep Equilibrium Modelscited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 35.

38 papers.

Sort:

By area:

  1. 2026A Mechanistic Analysis of Looped Reasoning Language Modelscited by 27
  2. 2026Recurrent-Depth VLA: Implicit Test-Time Compute Scaling of Vision-Language-Action Models via Latent Iterative Reasoningcited by 17
  3. 2025Universal Reasoning Modelcited by 11
  4. 2026Solve the Loop: Attractor Models for Language and Reasoningcited by 11
  5. 2025Tiny Recursive Models on ARC-AGI-1: Inductive Biases, Identity Conditioning, and Test-Time Computecited by 9
  6. 2026Are Your Reasoning Models Reasoning or Guessing? A Mechanistic Analysis of Hierarchical Reasoning Modelscited by 9
  7. 2026PRO-LONG: Programmatic Memory Enables Long-Horizon Reasoningcited by 9
  8. 2026Prioritize the Process, Not Just the Outcome: Rewarding Latent Thought Trajectories Improves Reasoning in Looped Language Modelscited by 6
  9. 2026Towards Agentic Intelligence for Materials Sciencecited by 3
  10. 2026SokoBench: Evaluating Long-Horizon Planning and Reasoning in Large Language Modelscited by 2
  11. 2026Test-Time Compute Scaling for ASR with Depth-Conditioned Looped Transformerscited by 2
  12. 2026What Makes Effective Supervision in Latent Chain-of-Thought: An Information-Theoretic Analysiscited by 2
26 more
  1. 2026Quantizing Recursive Reasoning Modelscited by 2
  2. 2026From Intents to Actions: Agentic AI in Autonomous Networkscited by 1
  3. 2026Thinking into the Future: Latent Lookahead Training for Transformerscited by 1
  4. 2026The Thinking Pixel: Recursive Sparse Reasoning in Multimodal Diffusion Latentscited by 1
  5. 2026Measuring AI Reasoning: A Guide for Researcherscited by 1
  6. 2026Formalizing Latent Thoughts: Four Axioms of Thought Representation in LLMscited by 1
  7. 2026Can Video World Models Track Unobserved World States?cited by 1
  8. 2025HRM-Agent: Training a recurrent reasoning model in dynamic environments using reinforcement learningcited by 0
  9. 2026Reasoning is a Modalitycited by 0
  10. 2026Grow-Prune-Freeze Networks: Adaptive & Continual Learning Technique for Olfactory Navigationcited by 0
  11. 2026Language-Guided Abstraction for Visual Reasoningcited by 0
  12. 2026DiARC: Distinguishing Positive and Negative Samples Helps Improving ARC-like Reasoning Ability of Large Language Modelscited by 0
  13. 2026Recursive Vision Language Models for General Symbolic Reasoningcited by 0
  14. 2026Latent Recurrent Thoughts: Recurrent Refinement of Proposed Latents for Reasoning with Frozen LLMscited by 0
  15. 2026Dissecting Hierarchical Reasoning Models: A Mechanistic Studycited by 0
  16. 2026LoopCD: Loop-wise Contrastive Decoding for Improving Reasoning in Looped Language Modelscited by 0
  17. 2026Not Too Hard, Not Too Easy: Learning from Intermediate States for LLM Structured Reasoningcited by 0
  18. 2026Principled Thoughts for Latent Recursive LLM Systemscited by 0
  19. 2026OverForge: Reasoning Through Strategies and Tactics Helps Cooperative Lifelong Adaptationcited by 0
  20. 2026What Limits Recursive Reasoning Models: Optimization, Architecture and Test-Time Scalingcited by 0
  21. 2026Unrolled Flow Models for Reasoningcited by 0
  22. 2026Think Before You Paint: Recursive Latent Reasoning for Diffusion Modelscited by 0
  23. 2026ResidualQuant: KV Cache Quantization for Looped Transformers with 2-Bit Residualscited by 0
  24. 2026Invariant-Measure Reasoners: Stable Representations for Latent Reasoningcited by 0
  25. 2026From Chain-of-Thought to Loops: Non-Autoregressive Latent Reasoning via Looped Transformerscited by 0
  26. 2026Scaling to Tens of Thousands of Test-Time Iterations with Loop-Native Attention Residualscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Form Follows Function: Recursive Stem Model
  2. 2026Probabilistic Tiny Recursive Model
  3. 2026Dynamical Systems Theory Behind a Hierarchical Reasoning Model
  4. 2026Energy-guided Recursive Model
  5. 2026Tiny Recursive Reasoning with Mamba-2 Attention Hybrid
  6. 2026Random Recursive Models
  7. 2026Generative Recursive Reasoning
  8. 2026Boosting Inference with Guided Reasoning: Stochastic Exploration for Recursive Models