Mechanistic Interpretability Benchmark

university

https://mib-bench.github.io

AI & ML interests

Principled evaluation of mechanistic interpretability methods.

Recent Activity

amueller updated a Space 4 days ago

mib-bench/leaderboard

amueller updated a model 4 days ago

mib-bench/mib-circuits-example

amueller updated a Space 6 days ago

mib-bench/leaderboard

View all activity

mib-bench 's datasets 7

mib-bench/ravel

Viewer • Updated May 31 • 117k • 34

mib-bench/arithmetic_subtraction

Viewer • Updated May 31 • 20.9k • 55

mib-bench/arithmetic_addition

Viewer • Updated May 31 • 40.4k • 195

mib-bench/ioi

Viewer • Updated May 29 • 21k • 1.53k

mib-bench/arc_easy

Viewer • Updated Jan 25 • 4.01k • 127

mib-bench/arc_challenge

Viewer • Updated Jan 25 • 2k • 58

mib-bench/copycolors_mcqa

Viewer • Updated Jan 16 • 1.89k • 666