From faad138d188feffaab4eb4b905fb7ce3cb7842f7 Mon Sep 17 00:00:00 2001 From: Sarah Schwettmann Date: Sun, 14 Apr 2024 02:53:40 -0400 Subject: [PATCH] Update index.html --- index.html | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/index.html b/index.html index 5418496..9276c7b 100644 --- a/index.html +++ b/index.html @@ -177,7 +177,7 @@

Generating Synthetic Test Im

Image editing

-

Maia can also call the edit_images tool which uses an text-based image editing module (Instruct Pix2Pix) to make image edits according to prompts written by MAIA. MAIA uses this tool to causally intervene on input space in order to test specific hypotheses about system behavior (e.g. whether the presence of a certain feature is required for the observed behavior!)


+

MAIA can also call the edit_images tool which uses an text-based image editing module (Instruct Pix2Pix) to make image edits according to prompts written by MAIA. MAIA uses this tool to causally intervene on input space in order to test specific hypotheses about system behavior (e.g. whether the presence of a certain feature is required for the observed behavior!)


Synthetic Exemplars @@ -191,7 +191,19 @@

Image editing

-

MAIA uses tools to design experiments on other systems

+

Using MAIA to remove spurious features

+
+

MAIA composes interpretability subroutines into python programs to answer user queries about a system. What kind of experiments does MAIA design? Below we highlight example usage of individual tools to run experiments on neurons inside common vision architectures (CLIP, ResNet, DINO). These are experimental excerpts intended to demonstrate tool use (often, MAIA runs many more experiments to reach its final conclusion!) For full experiment logs, check out our interactive [neuron viewer].

+
+
+
+
+
+ +
+
+
+

Using MAIA to reveal biases

MAIA composes interpretability subroutines into python programs to answer user queries about a system. What kind of experiments does MAIA design? Below we highlight example usage of individual tools to run experiments on neurons inside common vision architectures (CLIP, ResNet, DINO). These are experimental excerpts intended to demonstrate tool use (often, MAIA runs many more experiments to reach its final conclusion!) For full experiment logs, check out our interactive [neuron viewer].