arXiv Machine Learning
Jul 1
Fork-Think with Confidence
arXiv:2606. 31484v1 Announce Type: new Abstract: Parallel thinking has enjoyed great success for boosting LLM performance on reasoning tasks without the need for any re-training.
Second Thought: Reasoning in Parallel as LLM Agents Act and Observe
arXiv:2608. 13667v1 Announce Type: new Abstract: LLM agents in the ReAct paradigm alternate between reasoning, acting, and observing, but deliberate reasoning is confined to the Thought phase: while the agent serializes an action and waits for the environment, its reasoning is frozen.
Related stories
arXiv AI
Sep 16
State of Thought Enables Endogenous Reasoning
arXiv:2609.16055v1 Announce Type: cross Abstract: Test-time compute has emerged as a major approach to improving the capabilities of Large Language Models (LLMs). However, existing test-time reasonin...
arXiv AI
Aug 26
Parason: Revealing Subtask and Trial Parallelism in LLM Reasoning
Parason is a new framework that discovers and exploits both subtask and trial parallelism in large language model (LLM) reasoning. By converting sequential reasoning traces into structured parallel trajectories and training with Parallelism-Aware Group Relative Policy Optimization, it balances accuracy, latency, and parallelism. Experiments on mathematical reasoning benchmarks such as AIME24 and AIME25 show that Parason achieves an average acceleration of about 1.7× while maintaining competitive accuracy.
arXiv AI
Jun 2
ThinkSwitch: Context Distillation with LoRA and Weight Interpolation for Specific-Purpose Reasoning Tasks
arXiv:2606. 01080v1 Announce Type: cross Abstract: Large language models often improve on difficult tasks by spending inference-time compute on a reasoning trace before producing the final answer.
arXiv AI
Sep 4
</think> Doesn't Stop Reasoning: Analysis of Spurious CoT Termination
The paper investigates a training‑free early‑exit technique that inserts an end‑of‑think (EoT) token to terminate chain‑of‑thought (CoT) reasoning in large reasoning models. It finds that the injected EoT often fails to cleanly switch the model from reasoning to answering, leading to continued reasoning‑like generation—termed spurious CoT termination—whose length scales with the amount of reasoning saved. By increasing attention to the EoT token through Exit‑token Attention Biasing (EAB), the authors reduce spurious termination and shorten the answering phase across multiple models and benchmarks.
arXiv Machine Learning
4d ago
Act First, Reason Later: Accelerating On-Policy Distillation for Multi-Turn Agents via Reference-Conditioned Inverse Dynamics
arXiv:2609.36608v1 Announce Type: new Abstract: On-policy distillation (OPD) trains multi-turn language agents with dense teacher supervision on student-generated responses. However, standard think-t...
arXiv AI
Aug 18
ParaTempo: Efficient Parallel Reasoning via Temporal Confidence
arXiv:2608. 16425v1 Announce Type: new Abstract: Parallel reasoning improves the accuracy and robustness of large reasoning models by exploring multiple solution paths, but its computational cost grows with reasoning depth and branch count.
arXiv AI
Sep 2
Latent Recurrent Thoughts: Recurrent Refinement of Proposed Latents for Reasoning with Frozen LLMs
Latent Recurrent Thoughts (LRT) proposes a method for reasoning with frozen large language models by operating in the model’s continuous representation space. A small auxiliary network generates initial latent vectors, which a tiny recurrent reasoner refines over multiple steps, decoupling computational depth from model size. Experiments on symbolic and natural‑language reasoning tasks show that LRT outperforms prior frozen‑decoder continuous‑space methods and chain‑of‑thought prompting while using far less inference compute.
arXiv AI
Sep 1
Geometry of Divergence: Tracking Hidden-State Trajectories for Adaptive Multi-Turn Reasoning
arXiv:2608.30650v1 Announce Type: new Abstract: LLM agents need to sustain goal-consistent reasoning across long multi-turn interactions under strict resource constraints. However, as the multi-turn...
arXiv AI
Sep 3
When Can Large Reasoning Models Save Thinking? Mechanistic Analysis of Behavioral Divergence in Reasoning
The paper investigates why large reasoning models (LRMs) often continue to think even when prompted to stop, a phenomenon called "Still-thinking". By examining confidence at the thinking-termination boundary, internal attention divergences, and attention allocation across prompt segments, the authors find that high perplexity and greater attention to the original question correlate with continued thinking. They propose an attention‑intervention method that suppresses explicit reasoning, which reduces inefficiency but also lowers accuracy, underscoring a trade‑off between instruction compliance, inference speed, and correctness.
arXiv AI
Jun 3
Agentic Chain-of-Thought Steering for Efficient and Controllable LLM Reasoning
arXiv:2606. 03965v1 Announce Type: cross Abstract: Large language models improve final-answer accuracy through extended chain-of-thought reasoning, but often spend tokens inefficiently and offer little inference-time control.
arXiv Machine Learning
Jul 2
Message Passing Enables Efficient Reasoning
arXiv:2607. 01077v1 Announce Type: cross Abstract: While inference-time scaling has improved the reasoning abilities of large language models (LLMs), the need to generate long chains-of-thought (CoTs) is a computational bottleneck.