EurekaBench: Measuring Agentic Ability to Discover New Scientific Insights

arXiv:2610.00492 · cs.CL, cs.AI · Submitted 2026-09-30 · Read on arXiv

cs.CL, cs.AI

Submitted: 2026-09-30

Updated: 2026-09-30

Code: https://github.com/TransformerLensOrg/TransformerLens

Terminology

• AI Agents — 25× in this paper · explained in 2 episodes
• Accuracy — 25× in this paper · explained in 1 episode
• Benchmark — 15× in this paper · explained in 2 episodes
• alpha — 15× in this paper · explained in 1 episode
• Steering — 12× in this paper · explained in 2 episodes
• Trace — 9× in this paper · explained in 2 episodes
• Core — 7× in this paper · explained in 2 episodes
• Benchmarking — 6× in this paper · explained in 3 episodes
• Adaptation — 4× in this paper · explained in 1 episode
• Baryonic Feedback — 3× in this paper · explained in 3 episodes
• Abstraction — 2× in this paper · explained in 1 episode
• Adam — 2× in this paper · explained in 2 episodes
• Alternation — 2× in this paper · explained in 1 episode
• Calibration — 2× in this paper · explained in 7 episodes
• Interpretability — 2× in this paper · explained in 15 episodes
• Ablation — 1× in this paper · explained in 1 episode
• Additivity — 1× in this paper · explained in 1 episode
• Agentic Discovery — 1× in this paper · explained in 1 episode
• Computational Biology — 1× in this paper · explained in 2 episodes
• DFT Calculations — 1× in this paper · explained in 2 episodes
• Density Functional Theory (DFT) — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• Generalized Perturbed Equilibrium Code — 1× in this paper
• Grounding — 1× in this paper · explained in 3 episodes
• Inverse Problem — 1× in this paper · explained in 2 episodes
• Multi-Agent System — 1× in this paper · explained in 6 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Reinforcement Learning (RL) — 1× in this paper · explained in 25 episodes
• Reward Function — 1× in this paper · explained in 2 episodes
• World Model — 1× in this paper · explained in 7 episodes

Sources

• MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering
• PhysGym: Benchmarking LLMs in Interactive Physics Discovery with Controlled Priors
• ScienceAgentBench: Toward Rigorous Assessment of Language Agents for Data-Driven Scientific Discovery
• TORAX: A Fast and Differentiable Tokamak Transport Simulator in JAX
• Subliminal Learning: Language models transmit behavioral traits via hidden signals in data
• On the sensitivity of different galaxy properties to warm dark matter
• A foundation model of vision, audition, and language for in-silico neuroscience
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• Training AI Scientists to Replicate Research
• BoxingGym: Benchmarking Progress in Automated Experimental Design and Model Discovery
• The DREAMS Project: Disentangling the Impact of Halo-to-Halo Variance and Baryonic Feedback on Milky Way Dark Matter Density Profiles
• Are Large Language Models Reliable AI Scientists? Assessing Reverse-Engineering of Black-Box Systems
• Accumulating Context Changes the Beliefs of Language Models
• MaD Physics: Evaluating information seeking under constraints in physical environments
• DISCOVERYWORLD: A Virtual Environment for Developing and Evaluating Automated Scientific Discovery Agents
• LLM-AutoSciLab: Closed-Loop Scientific Discovery via Active Experimentation with LLMs
• On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists
• Gravity-Bench-v1: A Benchmark on Gravitational Physics Discovery for Agents
• EXP-Bench: Can AI Conduct AI Research Experiments?
• Linking Warm Dark Matter to Merger Tree Histories via Deep Learning Networks

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering