Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoning

• URL: http://arxiv.org/abs/2508.19202v1
• Date: Tue, 26 Aug 2025 17:04:23 GMT
• Title: Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoning
• Authors: Alan Li, Yixin Liu, Arpan Sarkar, Doug Downey, Arman Cohan,
• Abstract summary: We introduce SciReas, a diverse suite of existing benchmarks for scientific reasoning tasks.<n>We then propose KRUX, a probing framework for studying the distinct roles of reasoning and knowledge in scientific tasks.
• Score: 53.82037883518254
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Scientific problem solving poses unique challenges for LLMs, requiring both deep domain knowledge and the ability to apply such knowledge through complex reasoning. While automated scientific reasoners hold great promise for assisting human scientists, there is currently no widely adopted holistic benchmark for evaluating scientific reasoning, and few approaches systematically disentangle the distinct roles of knowledge and reasoning in these tasks. To address these gaps, we introduce SciReas, a diverse suite of existing benchmarks for scientific reasoning tasks, and SciReas-Pro, a selective subset that requires more complex reasoning. Our holistic evaluation surfaces insights about scientific reasoning performance that remain hidden when relying on individual benchmarks alone. We then propose KRUX, a probing framework for studying the distinct roles of reasoning and knowledge in scientific tasks. Combining the two, we conduct an in-depth analysis that yields several key findings: (1) Retrieving task-relevant knowledge from model parameters is a critical bottleneck for LLMs in scientific reasoning; (2) Reasoning models consistently benefit from external knowledge added in-context on top of the reasoning enhancement; (3) Enhancing verbalized reasoning improves LLMs' ability to surface task-relevant knowledge. Finally, we conduct a lightweight analysis, comparing our science-focused data composition with concurrent efforts on long CoT SFT, and release SciLit01, a strong 8B baseline for scientific reasoning.

Related papers

• Probing Scientific General Intelligence of LLMs with Scientist-Aligned Workflows [203.3527268311731]
  We present an operational SGI definition grounded in the Practical Inquiry Model (PIM)<n>We operationalize it via four scientist-aligned tasks: deep research, idea generation, dry/wet experiments, and experimental reasoning.<n>Our PIM-grounded definition, workflow-centric benchmark, and empirical insights establish a foundation for AI systems that genuinely participate in scientific discovery.
  arXiv  Detail & Related papers  (2025-12-18T12:44:36Z)
• Can AI Validate Science? Benchmarking LLMs for Accurate Scientific Claim $\ ightarrow$ Evidence Reasoning [6.043212666944194]
  We present CLAIM-BENCH, a benchmark for evaluating large language models' capabilities in scientific claim-evidence extraction and validation.<n>We show that closed-source models like GPT-4 and Claude consistently outperform open-source counterparts in precision and recall.<n> strategically designed three-pass and one-by-one prompting approaches significantly improve LLMs' abilities to accurately link dispersed evidence with claims.
  arXiv  Detail & Related papers  (2025-06-09T21:04:39Z)
• LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models [20.800445482814958]
  Large Language Models (LLMs) have gained interest for their potential to leverage embedded scientific knowledge for hypothesis generation.<n>Existing benchmarks often rely on common equations that are susceptible to memorization by LLMs, leading to inflated performance metrics that do not reflect discovery.<n>In this paper, we introduce LLM-SRBench, a comprehensive benchmark with 239 challenging problems across four scientific domains.<n>Our benchmark comprises two main categories: LSR-Transform, which transforms common physical models into less common mathematical representations to test reasoning beyond memorized forms, and LSR- Synth, which introduces synthetic, discovery-driven problems requiring data-driven reasoning
  arXiv  Detail & Related papers  (2025-04-14T17:00:13Z)
• ResearchBench: Benchmarking LLMs in Scientific Discovery via Inspiration-Based Task Decomposition [67.26124739345332]
  Large language models (LLMs) have demonstrated potential in assisting scientific research, yet their ability to discover high-quality research hypotheses remains unexamined.<n>We introduce the first large-scale benchmark for evaluating LLMs with a near-sufficient set of sub-tasks of scientific discovery.<n>We develop an automated framework that extracts critical components - research questions, background surveys, inspirations, and hypotheses - from scientific papers.
  arXiv  Detail & Related papers  (2025-03-27T08:09:15Z)
• Auto-Bench: An Automated Benchmark for Scientific Discovery in LLMs [23.608962459019278]
  We introduce a novel benchmark to evaluate Large Language Models (LLMs) for scientific discovery in both natural and social sciences.<n>Our benchmark is based on the principles of causal graph discovery. It challenges models to uncover hidden structures and make optimal decisions, which includes generating valid justifications.<n>We evaluate state-of-the-art LLMs, including GPT-4, Gemini, Qwen, Claude, and Llama, and observe a significant performance drop as the problem complexity increases.
  arXiv  Detail & Related papers  (2025-02-21T05:35:20Z)
• Position: Multimodal Large Language Models Can Significantly Advance Scientific Reasoning [51.11965014462375]
  Multimodal Large Language Models (MLLMs) integrate text, images, and other modalities.<n>This paper argues that MLLMs can significantly advance scientific reasoning across disciplines such as mathematics, physics, chemistry, and biology.
  arXiv  Detail & Related papers  (2025-02-05T04:05:27Z)
• SciLitLLM: How to Adapt LLMs for Scientific Literature Understanding [22.131371019641417]
  Despite Large Language Models' success, they face challenges in scientific literature understanding.<n>We propose a hybrid strategy that integrates continual pre-training (CPT) and supervised fine-tuning (SFT)<n>We present a suite of LLMs: SciLitLLM, specialized in scientific literature understanding.
  arXiv  Detail & Related papers  (2024-08-28T05:41:52Z)
• SciKnowEval: Evaluating Multi-level Scientific Knowledge of Large Language Models [35.98892300665275]
  We introduce the SciKnowEval benchmark, a framework that evaluates large language models (LLMs) across five progressive levels of scientific knowledge. These levels aim to assess the breadth and depth of scientific knowledge in LLMs, including memory, comprehension, reasoning, discernment, and application. We benchmark 26 advanced open-source and proprietary LLMs using zero-shot and few-shot prompting strategies.
  arXiv  Detail & Related papers  (2024-06-13T13:27:52Z)
• LLM and Simulation as Bilevel Optimizers: A New Paradigm to Advance Physical Scientific Discovery [141.39722070734737]
  We propose to enhance the knowledge-driven, abstract reasoning abilities of Large Language Models with the computational strength of simulations. We introduce Scientific Generative Agent (SGA), a bilevel optimization framework. We conduct experiments to demonstrate our framework's efficacy in law discovery and molecular design.
  arXiv  Detail & Related papers  (2024-05-16T03:04:10Z)
• SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models [70.5763210869525]
  We introduce an expansive benchmark suite SciBench for Large Language Model (LLM) SciBench contains a dataset featuring a range of collegiate-level scientific problems from mathematics, chemistry, and physics domains. The results reveal that the current LLMs fall short of delivering satisfactory performance, with the best overall score of merely 43.22%.
  arXiv  Detail & Related papers  (2023-07-20T07:01:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.