Latest AI for Science Research Papers

Research on applying AI to scientific discovery, drug design, materials science, and scientific computing.

4 Papers
Showing 4 of 4 papers

Learning to Discover at Test Time

Mert Yuksekgonul, Daniel Koceja, Xinhao Li +8 more

How can we use AI to discover a new state of the art for a scientific problem? Prior work in test-time scaling, such as AlphaEvolve, performs search by prompting a frozen LLM. We perform reinforcement learning at test time, so the LLM can continue to train, but now with experience specific to the te...

reinforcement learningtest-time trainingcontinual learningsearch subroutinelearning objective+2 more
Jan 22, 202635

Beyond Static Tools: Test-Time Tool Evolution for Scientific Reasoning

Jiaxuan Lu, Ziyu Kong, Yemin Wang +10 more

The central challenge of AI for Science is not reasoning alone, but the ability to create computational methods in an open-ended scientific world. Existing LLM-based agents rely on static, pre-defined tool libraries, a paradigm that fundamentally fails in scientific domains where tools are sparse, h...

LLM-based agentstool librariesscientific reasoningcomputational methodstest-time tool evolution+5 more
Jan 12, 202643
View all categories
Latest AI for Science Research | AI for Science Papers