London

Harry Mayne

Advanced
@HarryMayne

Hey! I'm a PhD@Oxford working on LLM explainability and building some evals along the way. You can check out all my research at www.harrymayne.com

qwen_3_chat_templates. Alternative chat templates for Qwen 3 8B. Useful for multi-turn RL

15

SV_interpretability. Code for the paper "Can sparse autoencoders be used to decompose and interpret steering vectors?"

8

faithfulness. Code for the paper "A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behavior"

6

SCEs. Code for the paper LLMs Don't Know Their Own Decision Boundaries: The Unreliability of Self-Generated Counterfactual Explanations

3

bookmark2email. Python

1

ICU-patient-subgroups. Unsupervised Learning Approaches for Identifying ICU Patient Subgroups: Do Results Generalise?

1
6
Apply