mlsanityData to 14 Oct 2026

A Model with No Head and Many Thoughts

arXiv 2608.31069 ↗First posted 31 Aug 2026cs.LG cs.CLcited by 0

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

Abstract (arXiv)

Large language models decode by projecting hidden states through a large vocabulary head at every step. This operation is computationally costly and forces all reasoning to be expressed in discrete tokens. We introduce Soft Latent Thinking, a method that replaces the LM head during reasoning with a lightweight projector, enabling autoregressive rollout in embedding space where reasoning steps remain continuous rather than tokenized. Experiments on DeepSeek-Qwen-1.5B and LLaMA-3.2-3B show that Soft Latent Thinking consistently improves pass@k across all k while reducing per-step compute during chain-of-thought. Our method achieves the highest pass@32 among all soft-thinking approaches, demonstrating that effective reasoning can be carried out in continuous space without discrete token generation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 18 of its 24 references that are in the map. First posted on arXiv 31 Aug 2026.

Builds on

18 of 24 references.

References · 18

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2024The Llama 3 Herd of Modelscited by 14,715
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  4. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  5. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  6. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  7. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  8. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  9. 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
  10. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,739
  11. 2023SGLang: Efficient Execution of Structured Language Model Programscited by 1,141
  12. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 797
6 more
  1. 2025OpenThoughts: Data Recipes for Reasoning Modelscited by 304
  2. 2025Soft Thinking: Unlocking the Reasoning Potential of LLMs in Continuous Concept Spacecited by 117
  3. 2025LLMs are Single-threaded Reasoners: Demystifying the Working Mechanism of Soft Thinkingcited by 38
  4. 2025Soft Tokens, Hard Truthscited by 35
  5. 2024Diffusion of Thoughts: Chain-of-Thought Reasoning in Diffusion Language Modelscited by 27
  6. 2025SofT-GRPO: Surpassing Discrete-Token LLM Reinforcement Learning via Gumbel-Reparameterized Soft-Thinking Policy Optimizationcited by 21

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Tyler: Typed Latent Reasoning for Language Models -- When to Think, What to Compute, and How Much to Allocate
  2. 2025Improving Latent Reasoning in LLMs via Soft Concept Mixing
  3. 2026Why Struggle with Continuous Latents? Interpretable Discrete Latent Reasoning via Rendered Compression
  4. 2026SeLaR: Selective Latent Reasoning in Large Language Models
  5. 2025Soft Reasoning: Navigating Solution Spaces in Large Language Models through Controlled Embedding Exploration
  6. 2026Latent Recurrent Thoughts: Recurrent Refinement of Proposed Latents for Reasoning with Frozen LLMs
  7. 2026CoLT: Teaching Multi-Modal Models to Think with Chain of Latent Thoughts
  8. 2026Latent Reasoning with Normalizing Flows