MechBench: Can AI Scientific Agents Discover Mechanisms Beyond Phenomenal Laws?
cs.AI, cs.LG, cs.SC
Submitted: 2026-09-28
Updated: 2026-09-28
Code: https://github.com/tsinghua-fib-lab/MechBench
Sources
- Can LLMs Discover Scientific Laws in Real and Parallel Worlds?
- Rethinking Symbolic Regression Datasets and Benchmarks for Scientific Discovery
- LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models
- Think like a Scientist: Physics-guided LLM Agent for Equation Discovery
- A-SR: Self-Evolving Agentic LLMs for Symbolic Regression via Hierarchical Coordination
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems