mlsanityData to 14 Oct 2026

Test-Time Scaling with Repeated Sampling Improves Multilingual Text Generation

arXiv 2505.21941 ↗First posted 28 May 2025cs.CLcited by 2

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

Abstract (arXiv)

Inference-time scaling via repeated sampling has shown promise in reasoning tasks, but its effectiveness in multilingual generation remains underexplored. We evaluate this approach using perplexity- and reward-based verifiers on two multilingual benchmarks: the Aya Evaluation Suite and m-ArenaHard. Our results show consistent quality improvements, with gains exceeding 35% in some cases. While perplexity-based scoring is effective for open-ended prompts, only reward-based verifiers improve performance on tasks requiring reasoning (e.g., math, code). Our results demonstrate the broader utility of repeated sampling for multilingual text generation and underscore the importance of selecting right verifiers for the task.

Key facts. Cited by 2 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 15 of its 18 references that are in the map. First posted on arXiv 28 May 2025.

Builds on

15 of 18 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025s1: Simple test-time scaling31 Jan
  2. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug
  3. 2024Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for Problem-Solving with Language Models1 Aug

Other references · 12

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  3. 2024Qwen2.5 Technical Reportcited by 4,942
  4. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 991
  5. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 500
  6. 2024RewardBench: Evaluating Reward Models for Language Modelingcited by 499
  7. 2024Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuningcited by 177
  8. 2024Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMscited by 142
  9. 2024Aya Expanse: Combining Research Breakthroughs for a New Multilingual Frontiercited by 129
  10. 2024Uncertainty-aware Reward Model: Teaching Reward Models to Know What is Unknowncited by 74
  11. 2024Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Samplingcited by 61
  12. 2024M-RewardBench: Evaluating Reward Models in Multilingual Settingscited by 50

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2026Think Deep, Not Just Long: Measuring LLM Reasoning Effort via Deep-Thinking Tokenscited by 22
  2. 2025When Life Gives You Samples: The Benefits of Scaling up Inference Compute for Multilingual LLMscited by 9

Similar, not cited

Close in wording, no citation either way

  1. 2025Multilingual Test-Time Scaling via Initial Thought Transfer
  2. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?
  3. 2026Sample Count Is Not Enough: Candidate-Generation Strategy Shapes the Energy and Performance of LLM Test-Time Scaling
  4. 2026Parallel Test-Time Scaling with Multi-Sequence Verifiers
  5. 2025The Art of Scaling Test-Time Compute for Large Language Models
  6. 2025Inference-Time Scaling for Complex Tasks: Where We Stand and What Lies Ahead
  7. 2025Sample Complexity and Representation Ability of Test-time Scaling Paradigms
  8. 2025Revisiting Test-Time Scaling: A Survey and a Diversity-Aware Method for Efficient Reasoning