Clone this repository at <script src="https://gist.github.com/yihua/66ad80224f8e50020025dc76d138dd0a.js"></script> Save yihua ...

GPT-6 Astra edited Qwen’s weights and raised MATH-500 from 7.9% to 48.1% in author-run tests. How weight-space programming works—and its limits.

This release provides the BREST-AI 16-bit, four-view model developed at the University of Warwick with NHS and academic collaborators. It estimates an examination-level three-year risk score from a ...

Comparative Analysis of CNN Architectures and Transfer Learning Techniques for Image Classification Using the Intel Image Classification Dataset An academic, highly reproducible deep learning ...

Code, candidate pools, cached LLM decisions and analysis scripts for the paper. Every number, table and figure of the paper can be regenerated offline from the released candidate pools with one ...