DiscoverPhysics: Benchmarking LLMs for Out-of-the-Box Scientific Thinking
stat.ML, cs.LG
Submitted: 2026-05-25
Updated: 2026-08-31
Code: https://github.com/ax-ml/jax
Terminology
Sources
- Is machine learning good or bad for the natural sciences?
- A Probabilistic Framework for LLM-Based Model Discovery
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
- GPQA: A Graduate-Level Google-Proof Q&A Benchmark
- On the Measure of Intelligence
- Think like a Scientist: Physics-guided LLM Agent for Equation Discovery
- DiscoveryBench: Towards Data-Driven Discovery with Large Language Models
- BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments
- BixBench: a Comprehensive Benchmark for LLM-based Agents in Computational Biology
- Auto-Discovery-Bench: Diagnosing Structured State Tracking in Oracle-Guided Discovery
- AI scientists produce results without reasoning scientifically
- NewtonBench: Benchmarking Generalizable Scientific Law Discovery in LLM Agents
- PhysGym: Benchmarking LLMs in Interactive Physics Discovery with Controlled Priors
- Gravity-Bench-v1: A Benchmark on Gravitational Physics Discovery for Agents
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- One Permutation Is All You Need: Fast, Deterministic Feature Importance and Model Stress-Testing
- Online Conformal Prediction for Non-Exchangeable Panel Data
- Deep Time-Series Forecasting in 10 Years: A Survey