Title: Interactive Scalable Discovery of Concepts, Vulnerabilities, and Evolutions in Deep Learning
Haekyu Park
School of Computational Science & Engineering
College of Computing
Georgia Institute of Technology
Date: Wednesday, October 11, 2023
Time: 1:00pm - 3:00pm ET (10:00am - 12:00pm PT)
Location (in-person): Coda 114 (first floor conference room; just walk in, no special access needed)
Location (remote): https://gatech.zoom.us/j/98145134253?pwd=ZFhzK29xYXMzUmZsMmN6b3JnM2ZSUT09
Committee
Dr. Duen Horng (Polo) Chau - Advisor, Georgia Tech, Computational Science & Engineering
Dr. Judy Hoffman - Georgia Tech, School of Interactive Computing
Dr. Callie Hao - Georgia Tech, School of Electrical and Computer Engineering
Dr. Chao Zhang - Georgia Tech, Computational Science & Engineering
Dr. Diyi Yang - Stanford University, Computer Science Department
Abstract
Deep Neural Networks (DNNs) are increasingly prevalent, but deciphering their functionality and acquired knowledge is challenging. This lack of clarity jeopardizes transparency, trust, and the ability to address issues during deployment, underscoring a pressing demand for deep learning interpretability. How can we efficiently summarize the concepts models learn? How do these concepts evolve during training? And when models are at risk from potential threats, how do we explain the models’ vulnerabilities?
We address these concerns with a human-centered approach, by developing novel systems to interpret learned concepts, their evolution, and potential vulnerabilities within deep learning. This thesis focuses on three key thrusts:
(1) Scalable Automatic Visual Summarization of Concepts. We develop NeuroCartography, an interactive system that scalably summarizes and visualizes concepts learned by a large-scale DNN, such as InceptionV1 trained with 1.2M images. A large-scale human evaluation with 244 participants shows that NeuroCartography discovers coherent, human-meaningful concepts.
(2) Insights to Reveal Model Vulnerabilities. We develop scalable interpretation techniques to visualize and identify internal elements in DNNs, which are susceptible to potential harms, aiming to understand how these defects lead to incorrect predictions. We develop first-of-its-kind interactive systems such as Bluff that visually compares the activation pathways for benign and attacked images in DNNs, and SkeletonVis that explains how attacks manipulate human joint detection in human action recognition models.
(3) Scalable Discovery of Concept Evolution During Training. Our first-of-its-kind ConceptEvo unified interpretation framework holistically reveals the inception and evolution of learned concepts and their relationships during training. ConceptEvo enables powerful new ways to monitor model training and discover training issues, addressing critical limitations of existing post-training interpretation research. A large-scale human evaluation with 260 participants demonstrates that ConceptEvo identifies concept evolutions that are both meaningful to humans and important for class predictions.
This thesis contributes to information visualization, deep learning, and crucially, their intersection. We have developed open-source interactive interfaces, scalable algorithms, and a unified framework for interpreting DNNs across different models. Our work impacts academia, industry, and the government. For example, our work has contributed to the DARPA GARD program (Garanteeing AI Robustness against Deception). Additionally, our work has been recognized through a J.P. Morgan AI PhD Fellowship and 2022 Rising Stars in IEEE EECS. NeuroCartography has been highlighted as a top visualization publication (top 1%) invited to SIGGRAPH.
Zoom Meeting Invitation
Topic: Haekyu Park's PhD Thesis Defense
Time: Oct 11, 2023 01:00 PM Eastern Time (US and Canada)
Join Zoom Meeting
https://gatech.zoom.us/j/98145134253?pwd=ZFhzK29xYXMzUmZsMmN6b3JnM2ZSUT09
Meeting ID: 981 4513 4253
Passcode: 900898
One tap mobile
+13126266799,,98145134253#,,,,*900898# US (Chicago)
+16469313860,,98145134253#,,,,*900898# US
Dial by your location
• +1 312 626 6799 US (Chicago)
• +1 646 931 3860 US
• +1 929 205 6099 US (New York)
• +1 301 715 8592 US (Washington DC)
• +1 305 224 1968 US
• +1 309 205 3325 US
• +1 719 359 4580 US
• +1 253 205 0468 US
• +1 253 215 8782 US (Tacoma)
• +1 346 248 7799 US (Houston)
• +1 360 209 5623 US
• +1 386 347 5053 US
• +1 507 473 4847 US
• +1 564 217 2000 US
• +1 669 444 9171 US
• +1 669 900 6833 US (San Jose)
• +1 689 278 1000 US
Meeting ID: 981 4513 4253
Passcode: 900898
Find your local number: https://gatech.zoom.us/u/aCkXCyX8r
Join by SIP
Join by H.323
• 162.255.37.11 (US West)
• 162.255.36.11 (US East)
• 115.114.131.7 (India Mumbai)
• 115.114.115.7 (India Hyderabad)
• 213.19.144.110 (Amsterdam Netherlands)
• 213.244.140.110 (Germany)
• 103.122.166.55 (Australia Sydney)
• 103.122.167.55 (Australia Melbourne)
• 149.137.40.110 (Singapore)
• 64.211.144.160 (Brazil)
• 149.137.68.253 (Mexico)
• 69.174.57.160 (Canada Toronto)
• 65.39.152.160 (Canada Vancouver)
• 207.226.132.110 (Japan Tokyo)
• 149.137.24.110 (Japan Osaka)
Meeting ID: 981 4513 4253
Passcode: 900898