mlsanityData to 14 Oct 2026

Test-time Scaling Techniques in Theoretical Physics -- A Comparison of Methods on the TPBench Dataset

arXiv 2506.20729 ↗First posted 25 Jun 2025cs.LG astro-ph.CO cs.AIcited by 8

On the map: LLM reasoning → Spatial and embodied reasoning

Abstract (arXiv)

Large language models (LLMs) have shown strong capabilities in complex reasoning, and test-time scaling techniques can enhance their performance with comparably low cost. Many of these methods have been developed and evaluated on mathematical reasoning benchmarks such as AIME. This paper investigates whether the lessons learned from these benchmarks generalize to the domain of advanced theoretical physics. We evaluate a range of common test-time scaling methods on the TPBench physics dataset and compare their effectiveness with results on AIME. To better leverage the structure of physics problems, we develop a novel, symbolic weak-verifier framework to improve parallel scaling results. Our empirical results demonstrate that this method significantly outperforms existing test-time scaling approaches on TPBench. We also evaluate our method on AIME, confirming its effectiveness in solving advanced mathematical problems. Our findings highlight the power of step-wise symbolic verification for tackling complex scientific problems.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 22 of its 24 references that are in the map. First posted on arXiv 25 Jun 2025.

Builds on

22 of 24 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025Theoretical Physics Benchmark (TPBench) -- a Dataset and Study of AI Reasoning Capabilities in Theoretical Physics19 Feb

Other references · 21

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  3. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
  4. 2025s1: Simple test-time scalingcited by 1,341
  5. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 1,017
  6. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 991
  7. 2025rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinkingcited by 313
  8. 2024Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for Problem-Solving with Language Modelscited by 290
  9. 2024FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AIcited by 243
  10. 2025Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMscited by 163
  11. 2024On the Self-Verification Limitations of Large Language Models on Reasoning and Planning Taskscited by 142
  12. 2025Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scalingcited by 136
9 more
  1. 2025T1: Advancing Language Model Reasoning through Reinforcement Learning and Inference Scalingcited by 77
  2. 2025ReasonFlux: Hierarchical LLM Reasoning via Scaling Thought Templatescited by 63
  3. 2025Atom of Thoughts for Markov LLM Test-Time Scalingcited by 54
  4. 2025Sample, Scrutinize and Scale: Effective Inference-Time Search by Scaling Verificationcited by 50
  5. 2025Think Twice: Enhancing LLM Reasoning by Scaling Multi-round Test-time Thinkingcited by 46
  6. 2025Step Back to Leap Forward: Self-Backtracking for Boosting Reasoning of Language Modelscited by 27
  7. 2025LADDER: Self-Improving LLMs Through Recursive Problem Decompositioncited by 25
  8. 2025Scaling Inference-Efficient Language Modelscited by 16
  9. 2025CoAT: Chain-of-Associated-Thoughts Framework for Enhancing Large Language Models Reasoningcited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 7.

8 papers.

  1. 2025A Survey of Scientific Large Language Models: From Data Foundations to Agent Frontierscited by 37
  2. 2026MadEvolve: Evolutionary Optimization of Cosmological Algorithms with Large Language Modelscited by 6
  3. 2026gwBenchmarks: Stress-Testing LLM Agents on High-Precision Gravitational Wave Astronomycited by 3
  4. 2026LLMs with in-context learning for Algorithmic Theoretical Physicscited by 2
  5. 2026Competing with AI Scientists: Agent-Driven Approach to Astrophysics Researchcited by 1
  6. 2026Fine-Tuning Small Reasoning Models for Quantum Field Theorycited by 1
  7. 2026Decoupled Physical Modeling and Execution for Physics Reasoningcited by 1
  8. 2026The Future of AI-assisted Physics: Discovering, Verifying and Organizing Results via Formalizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Are We Scaling the Right Thing? A System Perspective on Test-Time Scaling
  2. 2025Scaling over Scaling: Exploring Test-Time Scaling Plateau in Large Reasoning Models
  3. 2025SSR: Speculative Parallel Scaling Reasoning in Test-time
  4. 2025ABench-Physics: Benchmarking Physical Reasoning in LLMs via High-Difficulty and Dynamic Physics Problems
  5. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?
  6. 2025PhysicsEval: Inference-Time Techniques to Improve the Reasoning Proficiency of Large Language Models on Physics Problems
  7. 2025SCALE: Selective Resource Allocation for Overcoming Performance Bottlenecks in Mathematical Test-time Scaling
  8. 2025The Art of Scaling Test-Time Compute for Large Language Models