mlsanityData to 14 Oct 2026

SSR: Speculative Parallel Scaling Reasoning in Test-time

arXiv 2505.15340 ↗First posted 21 May 2025cs.LGcited by 4

On the map: LLM reasoning → Test-time compute and efficient thinking

also in 1 more
Abstract (arXiv)

Large language models (LLMs) have achieved impressive results on multi-step mathematical reasoning, yet at the cost of high computational overhead. This challenge is particularly acute for test-time scaling methods such as parallel decoding, which increase answer diversity but scale poorly in efficiency. To address this efficiency-accuracy trade-off, we propose SSR (Speculative Parallel Scaling Reasoning), a training-free framework that leverages a key insight: by introducing speculative decoding at the step level, we can accelerate reasoning without sacrificing correctness. SSR integrates two components: a Selective Parallel Module (SPM) that identifies a small set of promising reasoning strategies via model-internal scoring, and Step-level Speculative Decoding (SSD), which enables efficient draft-target collaboration for fine-grained reasoning acceleration. Experiments on three mathematical benchmarks-AIME 2024, MATH-500, and LiveMathBench - demonstrate that SSR achieves strong gains over baselines. For instance, on LiveMathBench, SSR improves pass@1 accuracy by 13.84% while reducing computation to 80.5% of the baseline FLOPs. On MATH-500, SSR reduces compute to only 30% with no loss in accuracy.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 19 of its 25 references that are in the map. First posted on arXiv 21 May 2025.

Builds on

19 of 25 references.

Closest ancestors · 8

Same direction, newest first

  1. 2025Accelerating Large Language Model Reasoning via Speculative Search3 May
  2. 2025Efficient Reasoning for LLMs through Speculative Chain-of-Thought27 Apr
  3. 2025SpecReason: Fast and Accurate Inference-Time Compute via Speculative Reasoning10 Apr
  4. 2025Harnessing the Reasoning Economy: A Survey of Efficient Reasoning for Large Language Models31 Mar
  5. 2025Accelerate Parallelizable Reasoning via Parallel Decoding within One Sequence26 Mar
  6. 2025Think Twice: Enhancing LLM Reasoning by Scaling Multi-round Test-time Thinking25 Mar
  7. 2025Dynamic Parallel Tree Search for Efficient LLM Reasoning22 Feb
  8. 2025Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling10 Feb

Other references · 11

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2020Scaling Laws for Neural Language Modelscited by 6,586
  3. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  4. 2023Accelerating Large Language Model Decoding with Speculative Samplingcited by 1,006
  5. 2024Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Headscited by 843
  6. 2025S*: Test Time Scaling for Code Generationcited by 98
  7. 2024FastDecode: High-Throughput GPU-Efficient LLM Serving using Heterogeneous Pipelinescited by 49
  8. 2025m1: Unleash the Potential of Test-Time Scaling for Medical Reasoning with Large Language Modelscited by 46
  9. 2024Block Verification Accelerates Speculative Decodingcited by 30
  10. 2025A Survey on Feedback-based Multi-step Reasoning for Large Language Models on Mathematicscited by 11
  11. 2023Revisiting Sparse Retrieval for Few-shot Entity Linkingcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 2.

4 papers.

  1. 2025A Survey on Parallel Reasoningcited by 11
  2. 2025DiffAdapt: Difficulty-Adaptive Reasoning for Token-Efficient LLM Inferencecited by 11
  3. 2026Unified Generation and Self-Verification for Vision-Language Models via Advantage Decoupled Preference Optimizationcited by 7
  4. 2026Tandem Reinforcement Learning with Verifiable Rewardscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Scaling Up, Speeding Up: A Benchmark of Speculative Decoding for Efficient LLM Test-Time Scaling
  2. 2026Taming Speculative Search for Test-Time Scaling in LLM Serving
  3. 2025ThreadWeaver: Adaptive Threading for Efficient Parallel Reasoning in Language Models
  4. 2025Are We Scaling the Right Thing? A System Perspective on Test-Time Scaling
  5. 2025Accelerated Test-Time Scaling with Model-Free Speculative Sampling
  6. 2025textttSPECS: Faster Test-Time Scaling through Speculative Drafts
  7. 2025Scaling over Scaling: Exploring Test-Time Scaling Plateau in Large Reasoning Models
  8. 2025Scaling Laws for Speculative Decoding