Pith. sign in

REVIEW 10 cited by

Let's Sample Step by Step: Adaptive-Consistency for Efficient Reasoning and Coding with LLMs

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 2305.11860 v2 pith:CRZUFWNJ submitted 2023-05-19 cs.CL

Let's Sample Step by Step: Adaptive-Consistency for Efficient Reasoning and Coding with LLMs

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

A popular approach for improving the correctness of output from large language models (LLMs) is Self-Consistency - poll the LLM multiple times and output the most frequent solution. Existing Self-Consistency techniques always generate a constant number of samples per question, where a better approach will be to non-uniformly distribute the available budget based on the amount of agreement in the samples generated so far. In response, we introduce Adaptive-Consistency, a cost-efficient, model-agnostic technique that dynamically adjusts the number of samples per question using a lightweight stopping criterion. Our experiments over 17 reasoning and code generation datasets and three LLMs demonstrate that Adaptive-Consistency reduces sample budget by up to 7.9 times with an average accuracy drop of less than 0.1%. Our code and data are available at https://www.sample-step-by-step.info

discussion (0)

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

Forward citations

Cited by 10 Pith papers

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

  1. VecFontLLM: Anchor-Guided Direct Synthesis of Chinese Vector Fonts

    cs.CV 2026-07 conditional novelty 7.0

    An anchor-guided MLLM generates complex Chinese vector glyphs from one or a few style exemplars, decoupling coarse layout from Bézier curve completion to improve structure and editability.

  2. TSRouter: Dynamic Modality-Model Selection for Time Series Reasoning

    cs.LG 2026-07 accept novelty 6.5

    A heterogeneous GNN that jointly routes each time-series query to the best modality–model pair under a user-chosen accuracy–cost trade-off, yielding large gains on four reasoning tasks.

  3. TSRouter: Dynamic Modality-Model Selection for Time Series Reasoning

    cs.LG 2026-07 conditional novelty 6.0

    A heterogeneous-graph router jointly selects the optimal modality (text, vision, or both) and model per time series query, beating prior routing baselines and generalizing to unseen models and tasks.

  4. Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time Scaling

    cs.CL 2026-06 unverdicted novelty 6.0

    RL-trained lightweight controller using answer statistics improves trade-offs among correctness, latency, and total samples in adaptive sampling for LLM test-time scaling.

  5. Dual-Dimensional Consistency: Balancing Budget and Quality in Adaptive Inference-Time Scaling

    cs.AI 2026-05 conditional novelty 6.0

    DDC couples confidence-weighted Bayesian early stopping with trend-aware pruning of reasoning paths, cutting inference tokens by over 10x on some benchmarks while roughly maintaining accuracy.

  6. Optimal Bayesian Stopping for Efficient Inference of Consistent LLM Answers

    stat.ML 2026-02 conditional novelty 6.0

    Tracking only the counts of the two most frequent LLM answers is asymptotically as sample-efficient as the full Bayesian posterior in an adaptive self-consistency procedure, cutting LLM calls by up to half.

  7. Scaling with Confidence: Calibrating Confidence of LLMs for Adaptive Test Time Scaling

    cs.AI 2026-07 unverdicted novelty 5.0

    C3RL is a new RL algorithm combining correctness, calibration, and reference accuracy rewards to improve LLM confidence calibration, enabling CAS to outperform majority voting with up to 12.33x lower inference cost.

  8. Dual-Dimensional Consistency: Balancing Budget and Quality in Adaptive Inference-Time Scaling

    cs.AI 2026-05 unverdicted novelty 5.0

    DDC reduces token consumption by over 10x in LLM reasoning while maintaining or exceeding baseline accuracy across five benchmarks via adaptive path quality filtering.

  9. Self-Consistency Is Losing Its Edge: Diminishing Returns and Rising Costs in Modern LLMs

    cs.AI 2025-11 unverdicted novelty 4.0

    Empirical evaluation on Gemini 2.5 models shows self-consistency yields only 0.4% gain on HotpotQA and 1.6% on MATH-500 across 20 samples while token costs scale linearly, with performance plateauing or declining at h...

  10. The Prompt Engineering Report Distilled: Quick Start Guide for Life Sciences

    cs.CL 2025-09 unverdicted novelty 3.0

    The paper reduces a broad set of prompt engineering techniques to six core approaches and applies them to life sciences use cases while addressing common LLM pitfalls.