Demystifying Hidden-State Recurrence: Switchable Latent Reasoning with On-Policy Reinforcement Learning

arXiv:2606.13106 · cs.LG, cs.CL · Submitted 2026-06-11 · Read on arXiv

cs.LG, cs.CL

Submitted: 2026-06-11

Updated: 2026-10-10

Code: https://github.com/LARK-AI-Lab/SWITCH

Terminology

• Latent Reasoning — 4× in this paper · explained in 2 episodes
• GRPO — 2× in this paper · explained in 6 episodes
• Reinforcement Learning — 2× in this paper · explained in 4 episodes
• GRPO Objective — 1× in this paper · explained in 2 episodes
• Reasoning Traces — 1× in this paper · explained in 3 episodes
• Reinforcement Learning (RL) — 1× in this paper · explained in 25 episodes
• Trace — 1× in this paper · explained in 2 episodes

Sources

• Eliciting Latent Predictions from Transformers with the Tuned Lens
• Reasoning in the Dark: Interleaved Vision-Text Reasoning in Latent Space
• Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs
• Training Verifiers to Solve Math Word Problems
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• LLM Latent Reasoning as Chain of Superposition
• Latent-GRPO: Group Relative Policy Optimization for Latent Reasoning
• From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Step
• Think before you speak: Training Language Models With Pause Tokens
• Training Large Language Models to Reason in a Continuous Latent Space
• How to use and interpret activation patching
• Measuring Mathematical Problem Solving With the MATH Dataset
• LoRA: Low-Rank Adaptation of Large Language Models
• Let's Verify Step by Step
• Let's Think Dot by Dot: Hidden Computation in Transformer Language Models
• Proximal Policy Optimization Algorithms
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• CODI: Compressing Chain-of-Thought into Continuous Space via Self-Distillation
• HybridFlow: A Flexible and Efficient RLHF Framework
• SwiReasoning: Switch-Thinking in Latent and Explicit for Pareto-Superior Reasoning LLMs

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks