Andrei Alexandru inwaves

Pinned Loading

1. Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.

   Python 18.2k 2.9k

2. High-level interpretability on language models fine-tuned to play chess.

   Python 1

3. Shallow neural networks interpolate like cubic/polyharmonic splines.

   Shell 1

4. Experimenting with sharpness-aware minimisation on noisily labelled images.

   Python 1

5. Emulating the UKESM1.0 climate model for precipitation, surface temperature and snow thickness using a Gaussian Process.

   Jupyter Notebook

6. Replicating invariant causal representation learning.

   Jupyter Notebook 2