Shared token budgets between visible chain-of-thought and answers create a coupling tax that makes non-thinking competitive on math benchmarks, with a truncation decomposition predicting the crossover and split budgets improving results.
hub
Reasoning models can be effective without thinking
22 Pith papers cite this work. Polarity classification is still indexing.
22
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
roles
background 3polarities
background 3representative citing papers
Language models produce overcomplete reasoning traces where on average 46% of steps can be removed while preserving the answer in 86% of cases, with necessity concentrated in the top three steps.
Post-Reasoning boosts LLM accuracy by reversing the usual answer-after-reasoning order, delivering mean relative gains of 17.37% across 117 model-benchmark pairs with zero extra cost.
SWE-Router introduces trajectory-conditioned value-based routing for LLM agents on SWE tasks, with a Bayes-optimality theorem and empirical cost savings while retaining most strong-model performance.
Dynamic Rollout Editing reduces overthinking in RL-trained LLMs by editing post-answer continuations in successful rollouts and preferring the edited versions within GRPO groups.
DyCon dynamically controls reasoning depth in LRMs by modeling evolving difficulty from step-level embeddings, reducing redundant steps across multiple benchmarks.
ThoughtFold applies introspective redundancy detection within correct CoT trajectories to create sub-trajectory spectra, then uses masked preference optimization to penalize redundant explorations, yielding 56% token reduction on DeepSeek-R1-Distill-Qwen-7B while preserving accuracy.
Stopping large reasoning models at the first correct reasoning prefix improves accuracy up to 21% by avoiding harmful overthinking that destabilizes correct trajectories.
CES applies conditional bidirectional entropy control on top of DAPO to improve accuracy and shorten responses on mathematical benchmarks for 7B and 1.5B LLMs.
MAFIG is a multi-agent framework that uses LLM agents and evaluators to generate reading comprehension items with significantly higher adherence to specified feature constraints than single-agent baselines.
PUMA detects reasoning-level semantic redundancy to enable early exit in chains of thought, achieving 26.2% average token reduction across five LRMs and five benchmarks while preserving accuracy and CoT quality.
HypEHR is a hyperbolic embedding model for EHR data that uses Lorentzian geometry and hierarchy-aware pretraining to answer clinical questions nearly as well as large language models but with much smaller size.
Step-GRPO internalizes dynamic early exit into reasoning models via step-structured optimization, Dynamic Truncated Rollout, and Step-Aware Relative Reward, delivering 32% token reduction on Qwen3-8B with no accuracy loss.
PreRL applies reward-driven updates to P(y) in pre-train space, uses Negative Sample Reinforcement to prune bad reasoning paths and boost reflection, and combines with standard RL in Dual Space RL to outperform baselines on reasoning tasks.
Post-training on reasoning tasks sparks the emergence of specialized attention heads that enable structured computation, with SFT adding stable heads while GRPO uses dynamic activation and pruning tied to reward signals, and controllable think models relying on compensatory heads instead of specific
AutoVLA unifies semantic reasoning and trajectory planning in one autoregressive VLA model for end-to-end autonomous driving by tokenizing trajectories into discrete actions and using GRPO reinforcement fine-tuning to adaptively reduce unnecessary reasoning.
LRMs exhibit complete accuracy collapse beyond certain puzzle complexities, with reasoning effort rising then declining, outperforming standard LLMs only on medium-complexity tasks.
TRACE aggregates answer consistency and confidence trajectory over multiple reasoning steps to decide when to halt inference, reducing token usage by 25-30% while keeping accuracy within 1-2% of full reasoning.
Self-aligned reward uses relative perplexity differences to encourage concise, query-specific reasoning in LLMs, yielding 4% accuracy gains and 30% lower inference cost when added to PPO or GRPO.
A survey organizing techniques to achieve efficient reasoning in LLMs by shortening chain-of-thought outputs.
DeepSight uses parallel latent feature prediction in BEV for long-horizon world modeling and adaptive text reasoning to reach state-of-the-art closed-loop performance on the Bench2drive benchmark.
citing papers explorer
-
The Coupling Tax: How Shared Token Budgets Undermine Visible Chain-of-Thought Under Fixed Output Limits
Shared token budgets between visible chain-of-thought and answers create a coupling tax that makes non-thinking competitive on math benchmarks, with a truncation decomposition predicting the crossover and split budgets improving results.
-
Uncovering the Representation Geometry of Minimal Cores in Overcomplete Reasoning Traces
Language models produce overcomplete reasoning traces where on average 46% of steps can be removed while preserving the answer in 86% of cases, with necessity concentrated in the top three steps.
-
Post Reasoning: Improving the Performance of Non-Thinking Models at No Cost
Post-Reasoning boosts LLM accuracy by reversing the usual answer-after-reasoning order, delivering mean relative gains of 17.37% across 117 model-benchmark pairs with zero extra cost.
-
SWE-Router: Routing in Multi-turn Agentic Software Engineering Tasks
SWE-Router introduces trajectory-conditioned value-based routing for LLM agents on SWE tasks, with a Bayes-optimality theorem and empirical cost savings while retaining most strong-model performance.
-
Dynamic Rollout Editing for Reducing Overthinking in RL-Trained Reasoning Models
Dynamic Rollout Editing reduces overthinking in RL-trained LLMs by editing post-answer continuations in successful rollouts and preferring the edited versions within GRPO groups.
-
DyCon: Dynamic Reasoning Control via Evolving Difficulty Modeling
DyCon dynamically controls reasoning depth in LRMs by modeling evolving difficulty from step-level embeddings, reducing redundant steps across multiple benchmarks.
-
ThoughtFold: Folding Reasoning Chains via Introspective Preference Learning
ThoughtFold applies introspective redundancy detection within correct CoT trajectories to create sub-trajectory spectra, then uses masked preference optimization to penalize redundant explorations, yielding 56% token reduction on DeepSeek-R1-Distill-Qwen-7B while preserving accuracy.
-
Thinking Past the Answer: Evaluating Harmful Overthinking in Large Reasoning Models
Stopping large reasoning models at the first correct reasoning prefix improves accuracy up to 21% by avoiding harmful overthinking that destabilizes correct trajectories.
-
Taming the Thinker: Conditional Entropy Shaping for Adaptive LLM Reasoning
CES applies conditional bidirectional entropy control on top of DAPO to improve accuracy and shorten responses on mathematical benchmarks for 7B and 1.5B LLMs.
-
A Multi-Agent Framework for Feature-Constrained Difficulty Control in Reading Comprehension Item Generation
MAFIG is a multi-agent framework that uses LLM agents and evaluators to generate reading comprehension items with significantly higher adherence to specified feature constraints than single-agent baselines.
-
Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models
PUMA detects reasoning-level semantic redundancy to enable early exit in chains of thought, achieving 26.2% average token reduction across five LRMs and five benchmarks while preserving accuracy and CoT quality.
-
HypEHR: Hyperbolic Modeling of Electronic Health Records for Efficient Question Answering
HypEHR is a hyperbolic embedding model for EHR data that uses Lorentzian geometry and hierarchy-aware pretraining to answer clinical questions nearly as well as large language models but with much smaller size.
-
Step-GRPO: Internalizing Dynamic Early Exit for Efficient Reasoning
Step-GRPO internalizes dynamic early exit into reasoning models via step-structured optimization, Dynamic Truncated Rollout, and Step-Aware Relative Reward, delivering 32% token reduction on Qwen3-8B with no accuracy loss.
-
From $P(y|x)$ to $P(y)$: Investigating Reinforcement Learning in Pre-train Space
PreRL applies reward-driven updates to P(y) in pre-train space, uses Negative Sample Reinforcement to prune bad reasoning paths and boost reflection, and combines with standard RL in Dual Space RL to outperform baselines on reasoning tasks.
-
Thinking Sparks!: Emergent Attention Heads in Reasoning Models During Post Training
Post-training on reasoning tasks sparks the emergence of specialized attention heads that enable structured computation, with SFT adding stable heads while GRPO uses dynamic activation and pruning tied to reward signals, and controllable think models relying on compensatory heads instead of specific
-
AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning
AutoVLA unifies semantic reasoning and trajectory planning in one autoregressive VLA model for end-to-end autonomous driving by tokenizing trajectories into discrete actions and using GRPO reinforcement fine-tuning to adaptively reduce unnecessary reasoning.
-
The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
LRMs exhibit complete accuracy collapse beyond certain puzzle complexities, with reasoning effort rising then declining, outperforming standard LLMs only on medium-complexity tasks.
-
Efficient Test-Time Scaling via Temporal Reasoning Aggregation
TRACE aggregates answer consistency and confidence trajectory over multiple reasoning steps to decide when to halt inference, reducing token usage by 25-30% while keeping accuracy within 1-2% of full reasoning.
-
Self-Aligned Reward: Towards Effective and Efficient Reasoners
Self-aligned reward uses relative perplexity differences to encourage concise, query-specific reasoning in LLMs, yielding 4% accuracy gains and 30% lower inference cost when added to PPO or GRPO.
-
Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models
A survey organizing techniques to achieve efficient reasoning in LLMs by shortening chain-of-thought outputs.
-
DeepSight: Long-Horizon World Modeling via Latent States Prediction for End-to-End Autonomous Driving
DeepSight uses parallel latent feature prediction in BEV for long-horizon world modeling and adaptive text reasoning to reach state-of-the-art closed-loop performance on the Bench2drive benchmark.
- Can Aha Moments Be Fake? Towards Quantifying Decorative and True Thinking in Chain-of-Thought