Matthew Ho

Advanced
@matt-seb-ho

PhD student at UCSD studying LM reasoning and meta learning

matt-seb-ho.github.io

WikiWhy. WikiWhy is a new benchmark for evaluating LLMs' ability to explain between cause-effect relationships. It is a QA dataset containing 9000+ "why" question-answer-rationale triplets.

49

arc_memo. Python

46

gfn_ntp. Jupyter Notebook

9

cse251b-nanogpt-contest-public. Python

3

llm_wrapper. Python

1

geophysics_agent. Python

1

mem2. Python

1
7
Apply