SciPredict: Can LLMs Predict the Outcomes of Scientific Experiments in Natural Sciences?

2026-04-12

AI Summary

This study investigates whether large language models (LLMs) can accurately predict outcomes of natural science experiments and reliably support scientific decision-making. The authors introduce SciPredict, a novel benchmark comprising 405 recent empirical tasks spanning 33 subfields across physics, biology, and chemistry, to systematically evaluate LLMs’ predictive accuracy and calibration—complemented by a human expert control experiment. Results reveal that leading LLMs achieve only 14–26% accuracy; although some models slightly surpass the average human expert performance (~20%), they remain far from practically useful in research contexts. Moreover, LLMs consistently exhibit poor calibration, demonstrating markedly weaker self-assessment of prediction reliability compared to human experts. This work establishes the first dedicated benchmark and analytical framework for evaluating scientific reasoning capabilities in LLMs.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Accelerating scientific discovery requires the identification of which experiments would yield the best outcomes before committing resources to costly physical validation. While existing benchmarks evaluate LLMs on scientific knowledge and reasoning, their ability to predict experimental outcomes - a task where AI could significantly exceed human capabilities - remains largely underexplored. We introduce SciPredict, a benchmark comprising 405 tasks derived from recent empirical studies in 33 specialized sub-fields of physics, biology, and chemistry. SciPredict addresses two critical questions: (a) can LLMs predict the outcome of scientific experiments with sufficient accuracy? and (b) can such predictions be reliably used in the scientific research process? Evaluations reveal fundamental limitations on both fronts. Model accuracies are 14-26% and human expert performance is $\approx$20%. Although some frontier models exceed human performance model accuracy is still far below what would enable reliable experimental guidance. Even within the limited performance, models fail to distinguish reliable predictions from unreliable ones, achieving only $\approx$20% accuracy regardless of their confidence or whether they judge outcomes as predictable without physical experimentation. Human experts, in contrast, demonstrate strong calibration: their accuracy increases from $\approx$5% to $\approx$80% as they deem outcomes more predictable without conducting the experiment. SciPredict establishes a rigorous framework demonstrating that superhuman performance in experimental science requires not just better predictions, but better awareness of prediction reliability. For reproducibility all our data and code are provided at https://github.com/scaleapi/scipredict

Problem

• scientific experiment prediction
• large language models
• outcome reliability
• scientific discovery
• experimental guidance

Innovation

• SciPredict
• experimental outcome prediction
• large language models
• prediction calibration
• scientific discovery