Pith. sign in

REVIEW 32 cited by

Chain of Draft: Thinking Faster by Writing Less

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

arxiv 2502.18600 v2 pith:262XTRWM submitted 2025-02-25 cs.CL

Chain of Draft: Thinking Faster by Writing Less

classification cs.CL
keywords reasoningtaskschaindraftintermediatellmsonlyreducing
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Large Language Models (LLMs) have demonstrated remarkable performance in solving complex reasoning tasks through mechanisms like Chain-of-Thought (CoT) prompting, which emphasizes verbose, step-by-step reasoning. However, humans typically employ a more efficient strategy: drafting concise intermediate thoughts that capture only essential information. In this work, we propose Chain of Draft (CoD), a novel paradigm inspired by human cognitive processes, where LLMs generate minimalistic yet informative intermediate reasoning outputs while solving tasks. By reducing verbosity and focusing on critical insights, CoD matches or surpasses CoT in accuracy while using as little as only 7.6% of the tokens, significantly reducing cost and latency across various reasoning tasks. Our code and data are available at https://github.com/sileix/chain-of-draft.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 32 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Selective Latent Thinking: Adaptive Compression of LLM Reasoning Chains

    cs.CL 2026-05 unverdicted novelty 7.0

    SLT selectively compresses reasoning spans via anticipation and gating, trained in three stages including RL, yielding 22.7% higher accuracy than uniform latent baselines at similar compression and 58.4% shorter chain...

  2. PANDO: Efficient Multimodal AI Agents via Online Skill Distillation

    cs.AI 2026-05 unverdicted novelty 7.0

    PANDO introduces an online skill-distillation method with a structured library, reflection, demotion, routing, compression, and cache-aware prompting that reaches 58.3% success on 910 VisualWebArena tasks using 58-61%...

  3. Respecting Self-Uncertainty in On-Policy Self-Distillation for Efficient LLM Reasoning

    cs.AI 2026-05 unverdicted novelty 7.0

    EGRSD and CL-EGRSD advance the accuracy-length frontier in LLM reasoning by entropy-guided weighting of token-level distillation signals from the teacher.

  4. Post Reasoning: Improving the Performance of Non-Thinking Models at No Cost

    cs.AI 2026-05 conditional novelty 7.0

    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.

  5. TrigReason: Trigger-Based Collaboration between Small and Large Reasoning Models

    cs.AI 2026-04 unverdicted novelty 7.0

    TrigReason matches large reasoning model accuracy on math and science benchmarks by delegating most steps to small models and intervening selectively on three triggers, cutting latency by 43.9% and cost by 73.3%.

  6. OS-Pruner: Pruning Chains-of-Thought of Reasoning Models via Optimal Stopping

    cs.AI 2026-07 conditional novelty 6.0

    A lightweight optimal-stopping policy on frozen reasoning LLMs cuts CoT length 20–60% with minimal accuracy loss by trading answer correctness against token cost via a tunable λ.

  7. The Harness Effect: How Orchestration Design Sets the Token Economics of Enterprise Agentic AI

    cs.AI 2026-07 conditional novelty 6.0

    Swapping only the orchestration layer around six AI models cut token cost per task 41% and latency 44% with quality at parity, making orchestration a bigger cost lever than model choice.

  8. When LLMs Develop Languages: Symbolic Communication for Efficient Multi-Agent Reasoning

    cs.AI 2026-06 unverdicted novelty 6.0

    CLSR lets LLM agents evolve and route symbolic languages that reduce generated tokens by 3-6x versus chain-of-thought while keeping accuracy on benchmarks.

  9. What LLMs explain is not what they believe: Evaluating explanation sufficiency under models' own input beliefs

    cs.LG 2026-06 unverdicted novelty 6.0

    Proposes SCSuff metric for evaluating LLM explanation sufficiency via model-generated alternative inputs, showing explanations are typically insufficient and predictable from hidden states.

  10. DyCon: Dynamic Reasoning Control via Evolving Difficulty Modeling

    cs.AI 2026-06 unverdicted novelty 6.0

    DyCon dynamically controls reasoning depth in LRMs by modeling evolving difficulty from step-level embeddings, reducing redundant steps across multiple benchmarks.

  11. Compress-Distill: Reasoning Trace Compression for Efficient Knowledge Distillation

    cs.LG 2026-06 unverdicted novelty 6.0

    Post-hoc model-based compression of reasoning traces cuts training tokens to 12-30% and speeds training 2-7.6x while retaining up to 96% of raw-trace accuracy, though raw traces remain superior at every scale.

  12. Unified Data Selection for LLM Reasoning

    cs.CL 2026-05 unverdicted novelty 6.0

    High-Entropy Sum (HES) selects high-quality reasoning data for LLMs by summing entropy of the top highest-entropy tokens, matching full-dataset performance with top 20% in SFT and outperforming baselines in RFT and RL.

  13. TCARD: Nearly Balanced Two-Level Designs with Treatment Cardinality Constraints with an Application to LLM Prompt Engineering

    stat.ME 2026-05 unverdicted novelty 6.0

    Proposes nearly balanced TCARDs that minimize the first two generalized word-length pattern components, defines Φ_BCD criterion linked to classical optimality, and constructs designs via coordinate exchange with simul...

  14. Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models

    cs.CL 2026-05 unverdicted novelty 6.0

    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.

  15. Nice Fold or Hero Call: Learning Budget-Efficient Thinking for Adaptive Reasoning

    cs.AI 2026-05 unverdicted novelty 6.0

    BET reduces reasoning tokens by about 55% on average while improving performance across benchmarks by learning to short-solve easy queries, fold early on unsolvable ones, and preserve budget for hard solvable queries.

  16. Implicit Compression Regularization: Concise Reasoning via Internal Shorter Distributions in RL Post-Training

    cs.AI 2026-05 unverdicted novelty 6.0

    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...

  17. CRISP: Compressed Reasoning via Iterative Self-Policy Distillation

    cs.LG 2026-03 conditional novelty 6.0

    CRISP achieves 57-59% token reduction on MATH-500 with 9-16 point accuracy gains on Qwen3 models via iterative self-distillation of concise reasoning behavior.

  18. CRISP: Compressed Reasoning via Iterative Self-Policy Distillation

    cs.LG 2026-03 conditional novelty 6.0

    On-policy reverse-KL self-distillation from a “be concise” teacher cuts reasoning length up to 56% while preserving or slightly improving math accuracy across several models.

  19. Efficient Reasoning on the Edge

    cs.LG 2026-03 accept novelty 5.5

    LoRA adapters, budget-forced GRPO, dynamic switching, parallel verification and FPTQuant enable practical chain-of-thought reasoning on quantized Qwen2.5-7B for edge devices.

  20. CAT: Confidence-Adaptive Thinking for Efficient Reasoning of Large Reasoning Models

    cs.CL 2026-07 unverdicted novelty 5.0

    CAT uses intrinsic confidence signals in preference optimization to adapt reasoning length in LRMs, outperforming uniform compression baselines on accuracy across benchmarks.

  21. LASER: Load-Aware Serving with Early-Exit for Reasoning LLMs at the Edge

    cs.DC 2026-06 unverdicted novelty 5.0

    LASER reduces edge LLM serving latency by 17-38% and improves SLO satisfaction by 3-6% via load-aware adaptive early-exit thresholds and difficulty-aware budget pre-allocation, with 1% average accuracy cost.

  22. SlimSearcher: Training Efficiency-Aware Web Agents via Adaptive Reward Gating

    cs.LG 2026-06 unverdicted novelty 5.0

    SlimSearcher reduces tool-call rounds by 17-58% on GAIA, BrowseComp and XBenchDeepSearch while maintaining accuracy via Pareto filtration in SFT and Adaptive Reward Gating in RL.

  23. Thinking Economically: A Hierarchical Framework for Adaptive-Complexity Reasoning in LLMs

    cs.CL 2026-05 unverdicted novelty 5.0

    HAB applies coarse-to-fine budgeting to LLM reasoning, predicting per-problem depth and learning intra-step token budgets via PPL comparisons and adaptive Pareto optimization, yielding higher accuracy and lower token ...

  24. Revisiting Anthropomorphic Reflection Markers in Large Language Model Reasoning

    cs.CL 2026-05 unverdicted novelty 5.0

    Suppressing anthropomorphic reflection markers via prompt and token interventions preserves or improves LLM reasoning performance on four benchmarks while models continue marker-free verification.

  25. Reasoning Compression with Mixed-Policy Distillation

    cs.AI 2026-05 unverdicted novelty 5.0

    Mixed-Policy Distillation transfers concise reasoning behavior from larger to smaller LLMs by having the teacher compress student-generated trajectories, cutting token usage up to 27% while raising benchmark scores.

  26. SAT: Balancing Reasoning Accuracy and Efficiency with Stepwise Adaptive Thinking

    cs.AI 2026-04 unverdicted novelty 5.0

    SAT reduces reasoning tokens by up to 40% across multiple large reasoning models and benchmarks by adaptively pruning steps based on difficulty while maintaining or improving accuracy.

  27. Early Stopping Chain-of-thoughts in Large Language Models

    cs.CL 2025-09 conditional novelty 5.0

    ES-CoT shortens LLM chain-of-thought generation by tracking runs of identical step answers after linguistic markers, cutting tokens 16% on average while keeping accuracy comparable to full CoT across six datasets and ...

  28. Pruning Long Chain-of-Thought of Large Reasoning Models via Small-Scale Preference Optimization

    cs.AI 2025-08 unverdicted novelty 5.0

    LCPO reduces average LRM output length by over 50% across benchmarks via targeted preference optimization while preserving reasoning performance.

  29. Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models

    cs.CL 2025-03 accept novelty 5.0

    A survey organizing techniques to achieve efficient reasoning in LLMs by shortening chain-of-thought outputs.

  30. The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes

    cs.CL 2026-06 unverdicted novelty 4.0

    A literature survey that introduces a taxonomy for LLM reasoning paradigms, analyzes methodological trends, and synthesizes failure modes from over 300 papers.

  31. Measuring AI Reasoning: A Guide for Researchers

    cs.AI 2026-05 unverdicted novelty 4.0

    Reasoning in language models should be measured by the faithfulness and validity of their multi-step search processes and intermediate traces, not final-answer accuracy.

  32. Token-Operations-Oriented Inference Optimization Techniques for Large Models

    cs.SE 2026-06 unverdicted novelty 3.0

    The paper introduces a four-layer technical architecture for token-operations-oriented inference optimization in large models and reviews key technologies and industry status at each layer.