REVIEW 30 cited by
Concise Thoughts: Impact of Output Length on LLM Reasoning and Cost
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
Signed reviews
read the original abstract
Today's large language models (LLMs) can solve challenging question-answering tasks, and prompt engineering techniques, such as chain-of-thought (CoT), have gained attention for enhancing the explanation and correctness of outputs. However, many models and techniques tend to produce excessively verbose and lengthy answers, leading to issues with both conciseness and generation time. To address this, this paper analyzes the impact of output lengths on LLM inference pipelines by introducing and proposing novel metrics to evaluate the \textit{correct conciseness} of a model and related prompting techniques. Then, we examine the impact of controlling output length through a refined prompt engineering strategy, Constrained-CoT (CCoT), which encourages the model to produce more concise outputs. To better understand the effects of such a prompt, we also introduce two additional scores for analyzing the conciseness, measured in terms of redundancy and information flow in generated answers. Experiments on pretrained LLMs and multiple datasets demonstrate the benefits of the proposed metrics and the effectiveness of CCoT across different models.
Forward citations
Cited by 30 Pith papers
-
Fewer Tokens, Smaller Cache: Reward-Coordinated Efficient Reasoning
ReCo coordinates KV-cache compression, reflection-token logit penalties, and confidence-based early stopping under one per-step process reward, reducing tokens and latency while largely preserving accuracy.
-
Shorter Reasoning, Earlier Answers? An Evaluation of Reasoning Interfaces
Matched-horizon comparisons show that a numeric/concision prompt shortens Qwen3 reasoning without consistent accuracy gains, while lower gpt-oss effort wins early mainly by finishing sooner, not by reasoning better.
-
NELSSA: A GPU-PNM Heterogeneous System for Mixed-Length LLM Serving via Length-based Request Placement
A heterogeneous GPU-PNM serving system that routes short prompts to GPUs and long prompts to near-memory sparse-attention devices, with runtime migration for growing contexts, reports large throughput and tail-latency...
-
Contrastive On-Policy Distillation
Contrasting a teacher's token scores under light- and heavy-thinking prompts yields a training signal that cuts reasoning length roughly in half without hurting accuracy.
-
R2-Router: A New Paradigm for LLM Routing with Reasoning
R2-Router jointly selects the LLM and an output-token budget, modeling each model as a quality-cost curve rather than a fixed point, and reports 4-5x cost savings on its new R2-Bench.
-
Are Large Reasoning Models Interruptible?
Interrupting large reasoning models or changing the problem mid-thought sharply degrades accuracy, with up to 60% drops and distinct failure modes: reasoning leakage, panic, and self-doubt.
-
Can Reasoning Path still be Effective as Input? Bridging Post-Reasoning to Chain-of-Thought Compression
UCoT compresses chain-of-thought reasoning by feeding a large language model a compact continuous 'upfront thought' generated by a small model, cutting output tokens roughly in half on math benchmarks while staying cl...
-
Wisdom of the Crowd, Without the Crowd: A Socratic LLM for Asynchronous Deliberation on Perspectivist Data
A Socratic LLM that questions annotators during labeling improved post-deliberation accuracy and confidence compared to a prior synchronous human-deliberation benchmark.
-
EAT: QoS-Aware Edge-Collaborative AIGC Task Scheduling via Attention-Guided Diffusion Reinforcement Learning
EAT, an attention-guided diffusion reinforcement learning scheduler for collaborative edge AIGC, reduces Stable Diffusion inference latency by 56-74% versus baselines while keeping CLIP quality nearly unchanged.
-
AALC: Large Language Model Efficient Reasoning via Adaptive Accuracy-Length Control
AALC, an accuracy-aware length reward, reduces reasoning-model response length by over 50% on math benchmarks while preserving or improving accuracy.
-
AdapThink: Adaptive Thinking Preferences for Reasoning Language Model
AdapThink is an RL post-training framework that adaptively reduces overthinking and underthinking in reasoning language models by rewarding confidence-appropriate reasoning depth and diverse training samples.
-
Evaluating the Use of LLMs for Documentation to Code Traceability
LLMs identify documentation-to-code trace links with F1 up to 80.4%, outperforming TF-IDF, BM25, and CodeBERT, but their explanations and chain reconstructions need human oversight.
-
QFFT, Question-Free Fine-Tuning for Adaptive Reasoning
Removing the question from fine-tuning data preserves a model's concise reasoning style while adding reflective long reasoning, halving token use without losing math accuracy.
-
The Overthinker's DIET: Cutting Token Calories with DIfficulty-AwarE Training
DIET adapts RL token penalties to estimated per-problem difficulty, cutting token use by roughly 40% on math benchmarks while slightly improving Pass@1.
-
Efficient Long CoT Reasoning in Small Language Models
Binary cutting with on-policy validation prunes redundant chain-of-thought steps in teacher traces, letting 7B models keep most long-CoT accuracy while generating fewer tokens.
-
Thinking Fast and Right: Balancing Accuracy and Reasoning Length with Adaptive Rewards
An adaptive length-penalty reward for RL-trained LLMs reduces reasoning length by over 50% with small accuracy loss by automatically tightening and relaxing the penalty based on the model's current accuracy.
-
An Empirical Study of LLM Reasoning Ability Under Strict Output Length Constraint
Under strict output token budgets, LLM rankings change: smaller instruction-tuned models can beat larger or reasoning-specialized models, and mid-sized models are latency-optimal.
-
When More is Less: Understanding Chain-of-Thought Length in LLMs
LLM accuracy follows an inverted U in chain-of-thought length, with an optimal length that grows with task difficulty and shrinks with model capability.
-
Rethinking Chain-of-Thought from the Perspective of Self-Training
A chain-of-thought framework that uses semantic entropy to select task-specific prompts and to stop iterative reasoning early, improving zero-shot accuracy by 4.2% on ten benchmarks.
-
INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge
INCLUDE is a multilingual benchmark of 197,243 exam questions from local sources that evaluates how well LLMs handle regional and cultural knowledge.
-
LenGuard-GPC: Length Guarding with Guided-Prompt Consistency for Spatial Reasoning Reinforce Learning
LenGuard-GPC adds a token-level KL consistency reward between standard and guided prompts, plus a staged length bonus, to GRPO training of Qwen3-VL-8B and reports better accuracy with shorter responses on multi-view s...
-
LISA: Linear-Indexed Sparse Attention for Efficient Long-Context Reasoning
LISA replaces standard attention with linear attention plus indexer-selected sparse attention, claiming O(nM) inference cost and accuracy gains, but the complexity analysis ignores the indexer's full-context scoring.
-
REST: Stress Testing Large Reasoning Models by Asking Multiple Problems at Once
Asking a reasoning model several problems at once reveals large accuracy drops and exposes differences that single-question benchmarks miss.
-
Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs
The survey's L1/L2 taxonomy and benchmark show that current reasoning models waste compute on easy problems and underthink hard ones, motivating more adaptive inference.
-
Prolonged Reasoning Is Not All You Need: Certainty-Based Adaptive Routing for Efficient LLM/MLLM Reasoning
CAR routes each query to either a short answer or full reasoning based on the perplexity of the model's draft answer, improving accuracy and cutting token use on VQA, KIE, and math/common sense benchmarks.
-
Incorporating Token Usage into Prompting Strategy Evaluation
Prompting strategies show sharply diminishing accuracy returns as token usage increases, and the paper formalizes this with Big-O_tok and Token Cost metrics.
-
Towards Concise and Adaptive Thinking in Large Reasoning Models: A Survey
A comprehensive review that categorizes methods for shortening and adaptively triggering chain-of-thought reasoning in large language models.
-
Generative AI Act II: Test Time Scaling Drives Cognition Engineering
Test-time scaling techniques such as long chain-of-thought, tree search, and self-correction define the paper's 'cognition engineering' paradigm, which it surveys, taxonomizes, and tutorials.
-
A Report on the llms evaluating the high school questions
Eight LLM APIs score 10% to 64% accuracy on a hand-picked Gaokao math set, with rankings that contradict the paper's own weighted scores.
-
A Survey of State of the Art Large Vision Language Models: Alignment, Benchmark, Evaluations and Challenges
A survey that catalogs large vision-language models, their alignment methods, benchmarks, and challenges, but is compromised by inconsistent counts and misclassified entries.
Discussion (0). Continue with ORCID to comment.