The Gradient: Perspectives on AI

Martin Wattenberg: ML Visualization and Interpretability

Nov 16, 2023

01:42:05

Snipd AI

Martin Wattenberg, a professor at Harvard and co-founder of Google Research's People + AI Research (PAIR) initiative, discusses his background in ML visualization, skepticism of neural networks in the 1980s, organization of information in graphics, progressive disclosure of complexity in interface design, evolutionary conversation interfaces, developing tools for model understanding, and creating trust in ML systems.

AI Summary

Highlights

AI Chapters

Episode notes

Podcast summary created with Snipd AI

Quick takeaways

Visualization plays a crucial role in understanding and interpreting machine learning models, providing insight, connection, and narrative.

Data visualization enhances understanding and engagement by organizing and presenting large amounts of data in a clear and visually appealing manner.

Visualization techniques reveal shared representations within machine translation models, suggesting the presence of a shared representation facilitating accurate translation across different languages.

Interpretability tools like TensorFlow.js and TKAV aim to increase trust by visualizing model decisions and limitations, helping users make informed decisions and calibrate their trust in the system.

Deep dives

Interpreting machine learning models through visualization

Visualization plays a crucial role in understanding and interpreting machine learning models. Martin Wautenberg, co-founder of Google's People Plus AI Research Team, emphasizes the importance of visualization in providing insight, connection, and narrative. Through his work on interpretability and visualization tools, he aims to spark new ways of thinking about interaction and interface design in the AI world.

Understanding language models and their internal representations

01:50

Considering Your Audience: Approaches to Presenting Complex Information

01:46

Shared representations and translation tasks in language models

01:57

Geometric properties of word embeddings: syntax and semantics

04:32

Introduction

4min

Data Visualization and Neural Networks

23min

Baby Names and Social Data Analysis: Exploring Popularity Patterns and Collective Analysis

3min

The Dynamics of Editing on Wikipedia

9min

Developing Tools for Model Understanding

6min

Birth and Development of TensorFlow.js

9min

Contributions and Thoughts on TKAV Paper

2min

Trust Creation in Machine Learning

19min

Opportunities for Dashboards and Interaction Design

28min

In episode 99 of The Gradient Podcast, Daniel Bashir speaks to Professor Martin Wattenberg.

Professor Wattenberg is a professor at Harvard and part-time member of Google Research’s People + AI Research (PAIR) initiative, which he co-founded. His work, with long-time collaborator Fernanda Viégas, focuses on making AI technology broadly accessible and reflective of human values. At Google, Professor Wattenberg, his team, and Professor Viégas have created end-user visualizations for products such as Search, YouTube, and Google Analytics. Note: Professor Wattenberg is recruiting PhD students through Harvard SEAS—info here.

Have suggestions for future podcast guests (or other feedback)? Let us know here or reach us at editor@thegradient.pub

Subscribe to The Gradient Podcast: Apple Podcasts | Spotify | Pocket Casts | RSSFollow The Gradient on Twitter

Outline:

* (00:00) Intro

* (03:30) Prof. Wattenberg’s background

* (04:40) Financial journalism at SmartMoney

* (05:35) Contact with the academic visualization world, IBM

* (07:30) Transition into visualizing ML

* (08:25) Skepticism of neural networks in the 1980s

* (09:45) Work at IBM

* (10:00) Multiple scales in information graphics, organization of information

* (13:55) How much information should a graphic display to whom?

* (17:00) Progressive disclosure of complexity in interface design

* (18:45) Visualization as a rhetorical process

* (20:45) Conversation Thumbnails for Large-Scale Discussions

* (21:35) Evolution of conversation interfaces—Slack, etc.

* (24:20) Path dependence — mutual influences between user behaviors and technology, takeaways for ML interface design

* (26:30) Baby Names and Social Data Analysis — patterns of interest in baby names

* (29:50) History Flow

* (30:05) Why investigate editing dynamics on Wikipedia?

* (32:06) Implications of editing patterns for design and governance

* (33:25) The value of visualizations in this work, issues with Wikipedia editing

* (34:45) Community moderation, bureaucracy

* (36:20) Consensus and guidelines

* (37:10) “Neutral” point of view as an organizing principle

* (38:30) Takeaways

* PAIR

* (39:15) Tools for model understanding and “understanding” ML systems

* (41:10) Intro to PAIR (at Google)

* (42:00) Unpacking the word “understanding” and use cases

* (43:00) Historical comparisons for AI development

* (44:55) The birth of TensorFlow.js

* (47:52) Democratization of ML

* (48:45) Visualizing translation — uncovering and telling a story behind the findings

* (52:10) Shared representations in LLMs and their facility at translation-like tasks

* (53:50) TCAV

* (55:30) Explainability and trust

* (59:10) Writing code with LMs and metaphors for using

* More recent research

* (1:01:05) The System Model and the User Model: Exploring AI Dashboard Design

* (1:10:05) OthelloGPT and world models, causality

* (1:14:10) Dashboards and interaction design—interfaces and core capabilities

* (1:18:07) Reactions to existing LLM interfaces

* (1:21:30) Visualizing and Measuring the Geometry of BERT

* (1:26:55) Note/Correction: The “Atlas of Meaning” Prof. Wattenberg mentions is called Context Atlas

* (1:28:20) Language model tasks and internal representations/geometry

* (1:29:30) LLMs as “next word predictors” — explaining systems to people

* (1:31:15) The Shape of Song

* (1:31:55) What does music look like?

* (1:35:00) Levels of abstraction, emergent complexity in music and language models

* (1:37:00) What Prof. Wattenberg hopes to see in ML and interaction design

* (1:41:18) Outro

Links:

* Professor Wattenberg’s homepage and Twitter

* Harvard SEAS application info — Professor Wattenberg is recruiting students!

* Research

* Earlier work

* A Fuzzy Commitment Scheme

* Stacked Graphs—Geometry & Aesthetics

* A Multi-Scale Model of Perceptual Organization in Information Graphics

* Conversation Thumbnails for Large-Scale Discussions

* Baby Names and Social Data Analysis

* History Flow (paper)

* At Harvard and Google / PAIR

* Tools for Model Understanding: Facets, SmoothGrad, Attacking discrimination with smarter ML

* TensorFlow.js

* Visualizing translation

* TCAV

* Other ML papers:

* The System Model and the User Model: Exploring AI Dashboard Design (recent speculative essay)

* Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task

* Visualizing and Measuring the Geometry of BERT

* Artwork

* The Shape of Song

Get full access to The Gradient at thegradientpub.substack.com/subscribe

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

The Gradient: Perspectives on AI

Martin Wattenberg: ML Visualization and Interpretability

Podcast summary created with Snipd AI

Quick takeaways

Deep dives

Interpreting machine learning models through visualization

The power of data visualization in journalism

Uncovering shared representations in machine learning models

Designing interpretability tools for machine learning

The Challenge of Understanding Machine Learning

The System Model and the User Model

Visualizing and Measuring the Geometry of Language Models

Understanding language models and their internal representations

Considering Your Audience: Approaches to Presenting Complex Information

Shared representations and translation tasks in language models

Geometric properties of word embeddings: syntax and semantics

Get the Snipd
podcast app

AI-powered
podcast player

Discover
highlights

Save any
moment

Share
& Export

AI-powered
podcast player

Discover
highlights

The Gradient: Perspectives on AI

Martin Wattenberg: ML Visualization and Interpretability

Podcast summary created with Snipd AI

Quick takeaways

Deep dives

Interpreting machine learning models through visualization

The power of data visualization in journalism

Uncovering shared representations in machine learning models

Designing interpretability tools for machine learning

The Challenge of Understanding Machine Learning

The System Model and the User Model

Visualizing and Measuring the Geometry of Language Models

Understanding language models and their internal representations

Considering Your Audience: Approaches to Presenting Complex Information

Shared representations and translation tasks in language models

Geometric properties of word embeddings: syntax and semantics

Get the Snipdpodcast app

AI-poweredpodcast player

Discoverhighlights

Save anymoment

Share& Export

AI-poweredpodcast player

Discoverhighlights

Get the Snipd
podcast app

AI-powered
podcast player

Discover
highlights

Save any
moment

Share
& Export

AI-powered
podcast player

Discover
highlights