REVIEW 27 cited by
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Between Underthinking and Overthinking: An Empirical Study of Reasoning Length and correctness in LLMs
read the original abstract
Large language models (LLMs) are increasingly optimized for long reasoning, under the assumption that more reasoning leads to better performance. However, emerging evidence suggests that longer responses can sometimes degrade accuracy rather than improve it. In this paper, we conduct a systematic empirical study of the relationship between reasoning length and answer correctness. We find that LLMs tend to overthink simple problems, generating unnecessarily long outputs, and underthink harder ones, failing to extend their reasoning when it is most needed. This indicates that models might misjudge problem difficulty and fail to calibrate their response length appropriately. Furthermore, we investigate the effects of length reduction with a preference optimization algorithm when simply preferring the shorter responses regardless of answer correctness. Experiments show that the generation length can be significantly reduced while maintaining acceptable accuracy. Our findings highlight generation length as a meaningful signal for reasoning behavior and motivate further exploration into LLMs' self-awareness in reasoning length adaptation.
Forward citations
Cited by 27 Pith papers
-
Know When to Stop: Segment-Level Credit Assignment for Reducing Overthinking
DASH assigns segment-level credit in reasoning traces using drift toward ground-truth answers, yielding 50.8% accuracy on AIME25 versus 45.4% for GRPO while reducing overthinking behaviors.
-
Faster Synchronous On-Policy RL via Straggler-Aware Group Sizing
SAGC dynamically adjusts group sizes in synchronous GRPO and DAPO via online constrained optimization to cut stragglers, improve wall-clock speed, and maintain or improve rewards and downstream reasoning performance.
-
On the Cost and Benefit of Chain of Thought: A Learning-Theoretic Perspective
Chain of Thought risk decomposes into oracle-trajectory benefit and trajectory-mismatch cost, with stability determining bounded, linear, or exponential error growth.
-
GRASP: Learning to Ground Social Reasoning in Multi-Person Non-Verbal Interactions
GRASP is a large-scale dataset and benchmark for social reasoning grounded in gaze and gesture events in multi-person videos, with Social Grounding Reward (SGR) proposed to improve model performance on GRASP-Bench.
-
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.
-
Mitigating Factual Hallucination in Large Reasoning Models via Mixed-Mode Advantage Regularization
MARGO mitigates thinking-induced hallucination in large reasoning models by using mixed-mode GRPO rollout groups that compare thinking trajectories against same-model non-thinking references.
-
CARE: Competence-Aware Reward Shaping for Adaptive Reasoning Length in Video-MLLMs
CARE uses exponential moving average competence estimates to progressively shift RL rewards from exploration-oriented long reasoning to efficiency-oriented concise reasoning in video-MLLMs, with batch normalization an...
-
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.
-
An Asymptotic Theory of Chain-of-Thought in In-Context Learning
Exact RMT-derived formula for CoT generalization error in linear ICL reveals phase transition between exponential/polynomial improvement, saturation, and overthinking regimes depending on depth, pretraining, and conte...
-
Quantized Reasoning Models Think They Need to Think Longer, but They Do Not
Post-training quantization increases overthinking errors in reasoning models; a logit penalty on curated overthinking markers reduces CoT length 12-23% without accuracy loss.
-
LACO: Adaptive Latent Communication for Collaborative Driving
LACO introduces Iterative Latent Deliberation, Cross-Horizon Saliency Attribution, and Structured Semantic Knowledge Distillation to enable low-latency latent communication in collaborative driving while preserving pe...
-
CLORE: Content-Level Optimization for Reasoning Efficiency
CLORE augments correct on-policy rollouts by deleting repetitive and irrelevant segments then optimizes with auxiliary DPO to improve accuracy-efficiency trade-off on math benchmarks.
-
Efficient Agentic Reasoning Through Self-Regulated Simulative Planning
SR²AM achieves competitive Pass@1 accuracy on diverse tasks with 25.8-95.3% fewer reasoning tokens than much larger models by using self-regulated simulative planning trained via supervised learning and RL.
-
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.
-
Reasoning Models Don't Just Think Longer, They Move Differently
After length correction, reasoning-trained language models exhibit distinct hidden-state trajectory geometries on harder problems compared to instruction-tuned baselines, with the strongest effect in code domains.
-
Implicit Compression Regularization: Concise Reasoning via Internal Shorter Distributions in RL Post-Training
ICR creates a virtual shorter distribution from shortest correct on-policy responses to regularize RL post-training toward concise yet accurate reasoning, improving the accuracy-length Pareto frontier on math and know...
-
Towards Efficient Large Language Reasoning Models via Extreme-Ratio Chain-of-Thought Compression
Extra-CoT trains a semantic compressor on math CoT data, applies mixed-ratio SFT, and uses CHRPO reinforcement learning to achieve over 73% token reduction on MATH-500 with 0.6% accuracy gain on Qwen3-1.7B.
-
The Shape of Reasoning: Topological Analysis of Reasoning Traces in Large Language Models
A TDA-based framework is proposed to assess LLM reasoning trace quality, showing that topological features predict quality better than standard graph metrics.
-
Geometric Signatures of Reasoning: A Spectral Perspective on Task Hardness
Introduces effective dimension d_ρ from spectral analysis of reasoning trajectories to distinguish task hardness (0.93 AUC on MATH500) and uses kinematic features for early correctness prediction from partial generations.
-
Modularized Reinforcement Learning on LLMs: From MDP Creation to Exploration and Learning
Survey mapping RL techniques onto LLM training and highlighting gaps in value-based, off-policy, and bootstrapping methods.
-
Trust Region On-Policy Distillation
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
-
When Self-Belief Misleads: Active Label Acquisition for Reinforcement Learning with Verifiable Rewards
RLAVR uses the Corrective Advantage Gap metric and CARE policy to actively acquire ground-truth labels for key samples, stabilizing RLVR training and boosting performance with limited annotation budgets.
-
Reasoning Models Don't Just Think Longer, They Move Differently
After length-correcting hidden-state trajectories during chain-of-thought, reasoning models show systematically different geometry on harder problems than baselines, strongest in competitive programming.
-
Investigating Thinking Behaviours of Reasoning-Based Language Models for Social Bias Mitigation
Reasoning LLMs aggregate social biases through stereotype repetition and irrelevant information injection in their thinking processes, and a self-review prompt mitigates this on BBQ, StereoSet, and BOLD benchmarks.
-
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.
-
Critique of Agent Model
Distinguishes agentic (externally scaffolded) from agentive (internally structured) AI systems and proposes the Goal-Identity-Configurator architecture for endogenous autonomy.
-
Human vs Machine Mathematical Difficulty on Project Euler: An Experimental Analysis
Empirical study of frontier AI on Project Euler finds power-law machine effort scaling with human difficulty (b<1 for 20/25 models) and moderate support for exponential success probability decay, with SOTA 50% horizon...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.