December 01, 2025

“A Pragmatic Vision for Interpretability” by Neel Nanda

1 hour 3 minutes

Executive Summary

The Google DeepMind mechanistic interpretability team has made a strategic pivot over the past year, from ambitious reverse-engineering to a focus on pragmatic interpretability:

Trying to directly solve problems on the critical path to AGI going well

[[1]]

Carefully choosing problems according to our comparative advantage

Measuring progress with empirical feedback on proxy tasks

We believe that, on the margin, more researchers who share our goals should take a pragmatic approach to interpretability, both in industry and academia, and we call on people to join us

Our proposed scope is broad and includes much non-mech interp work, but we see this as the natural approach for mech interp researchers to have impact

Specifically, we’ve found that the skills, tools and tastes of mech interp researchers transfer well to important and neglected problems outside “classic” mech interp

See our companion piece for more on which research areas and theories of change we think are promising

Why pivot now? We think that times have changed.

Models are far more capable, bringing new questions within empirical reach

We have been [...]

---

Outline:

(00:10) Executive Summary

(03:00) Introduction

(03:44) Motivating Example: Steering Against Evaluation Awareness

(06:21) Our Core Process

(08:20) Which Beliefs Are Load-Bearing?

(10:25) Is This Really Mech Interp?

(11:27) Our Comparative Advantage

(14:57) Why Pivot?

(15:20) Whats Changed In AI?

(16:08) Reflections On The Fields Progress

(18:18) Task Focused: The Importance Of Proxy Tasks

(18:52) Case Study: Sparse Autoencoders

(21:35) Ensure They Are Good Proxies

(23:11) Proxy Tasks Can Be About Understanding

(24:49) Types Of Projects: What Drives Research Decisions

(25:18) Focused Projects

(28:31) Exploratory Projects

(28:35) Curiosity Is A Double-Edged Sword

(30:56) Starting In A Robustly Useful Setting

(34:45) Time-Boxing

(36:27) Worked Examples

(39:15) Blending The Two: Tentative Proxy Tasks

(41:23) What's Your Contribution?

(43:08) Jack Lindsey's Approach

(45:44) Method Minimalism

(46:12) Case Study: Shutdown Resistance

(48:28) Try The Easy Methods First

(50:02) When Should We Develop New Methods?

(51:36) Call To Action

(53:04) Acknowledgments

(54:02) Appendix: Common Objections

(54:08) Aren't You Optimizing For Quick Wins Over Breakthroughs?

(56:34) What If AGI Is Fundamentally Different?

(57:30) I Care About Scientific Beauty and Making AGI Go Well

(58:09) Is This Just Applied Interpretability?

(58:44) Are You Saying This Because You Need To Prove Yourself Useful To Google?

(59:10) Does This Really Apply To People Outside AGI Companies?

(59:40) Aren't You Just Giving Up?

(01:00:04) Is Ambitious Reverse-engineering Actually Overcrowded?

(01:00:48) Appendix: Defining Mechanistic Interpretability

(01:01:44) Moving Toward Mechanistic OR Interpretability

The original text contained 47 footnotes which were omitted from this narration.

---

First published:

December 1st, 2025

Source:

https://www.lesswrong.com/posts/StENzDcD3kpfGJssR/a-pragmatic-vision-for-interpretability

---

Narrated by TYPE III AUDIO.

...more

View all episodes

By LessWrong

December 01, 2025

“A Pragmatic Vision for Interpretability” by Neel Nanda

1 hour 3 minutes

Executive Summary

The Google DeepMind mechanistic interpretability team has made a strategic pivot over the past year, from ambitious reverse-engineering to a focus on pragmatic interpretability:

Trying to directly solve problems on the critical path to AGI going well

[[1]]

Carefully choosing problems according to our comparative advantage

Measuring progress with empirical feedback on proxy tasks

We believe that, on the margin, more researchers who share our goals should take a pragmatic approach to interpretability, both in industry and academia, and we call on people to join us

Our proposed scope is broad and includes much non-mech interp work, but we see this as the natural approach for mech interp researchers to have impact

Specifically, we’ve found that the skills, tools and tastes of mech interp researchers transfer well to important and neglected problems outside “classic” mech interp

See our companion piece for more on which research areas and theories of change we think are promising

Why pivot now? We think that times have changed.

Models are far more capable, bringing new questions within empirical reach

We have been [...]

---

Outline:

(00:10) Executive Summary

(03:00) Introduction

(03:44) Motivating Example: Steering Against Evaluation Awareness

(06:21) Our Core Process

(08:20) Which Beliefs Are Load-Bearing?

(10:25) Is This Really Mech Interp?

(11:27) Our Comparative Advantage

(14:57) Why Pivot?

(15:20) Whats Changed In AI?

(16:08) Reflections On The Fields Progress

(18:18) Task Focused: The Importance Of Proxy Tasks

(18:52) Case Study: Sparse Autoencoders

(21:35) Ensure They Are Good Proxies

(23:11) Proxy Tasks Can Be About Understanding

(24:49) Types Of Projects: What Drives Research Decisions

(25:18) Focused Projects

(28:31) Exploratory Projects

(28:35) Curiosity Is A Double-Edged Sword

(30:56) Starting In A Robustly Useful Setting

(34:45) Time-Boxing

(36:27) Worked Examples

(39:15) Blending The Two: Tentative Proxy Tasks

(41:23) What's Your Contribution?

(43:08) Jack Lindsey's Approach

(45:44) Method Minimalism

(46:12) Case Study: Shutdown Resistance

(48:28) Try The Easy Methods First

(50:02) When Should We Develop New Methods?

(51:36) Call To Action

(53:04) Acknowledgments

(54:02) Appendix: Common Objections

(54:08) Aren't You Optimizing For Quick Wins Over Breakthroughs?

(56:34) What If AGI Is Fundamentally Different?

(57:30) I Care About Scientific Beauty and Making AGI Go Well

(58:09) Is This Just Applied Interpretability?

(58:44) Are You Saying This Because You Need To Prove Yourself Useful To Google?

(59:10) Does This Really Apply To People Outside AGI Companies?

(59:40) Aren't You Just Giving Up?

(01:00:04) Is Ambitious Reverse-engineering Actually Overcrowded?

(01:00:48) Appendix: Defining Mechanistic Interpretability

(01:01:44) Moving Toward Mechanistic OR Interpretability

The original text contained 47 footnotes which were omitted from this narration.

---

First published:

December 1st, 2025

Source:

https://www.lesswrong.com/posts/StENzDcD3kpfGJssR/a-pragmatic-vision-for-interpretability

---

Narrated by TYPE III AUDIO.

...more

More shows like LessWrong (30+ Karma)

View all

Making Sense with Sam Harris

26,357 Listeners

Conversations with Tyler

2,456 Listeners

The Peter Attia Drive

8,665 Listeners

Sean Carroll's Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas

4,176 Listeners

ManifoldOne

93 Listeners

Your Undivided Attention

1,600 Listeners

All-In with Chamath, Jason, Sacks & Friedberg

9,909 Listeners

Machine Learning Street Talk (MLST)

93 Listeners

Dwarkesh Podcast

506 Listeners

Hard Fork

5,527 Listeners

The Ezra Klein Show

15,987 Listeners

Moonshots with Peter Diamandis

543 Listeners

No Priors: Artificial Intelligence | Technology | Startups

136 Listeners

Latent Space: The AI Engineer Podcast

95 Listeners

BG2Pod with Brad Gerstner and Bill Gurley

474 Listeners

Share “A Pragmatic Vision for Interpretability” by Neel Nanda

Sign up to save your podcasts

“A Pragmatic Vision for Interpretability” by Neel Nanda

“A Pragmatic Vision for Interpretability” by Neel Nanda

More shows like LessWrong (30+ Karma)

Making Sense with Sam Harris

Conversations with Tyler

The Peter Attia Drive

Sean Carroll's Mindscape: Science, Society, Philosophy, Culture, Arts, and Ideas

ManifoldOne

Your Undivided Attention

All-In with Chamath, Jason, Sacks & Friedberg

Machine Learning Street Talk (MLST)

Dwarkesh Podcast

Hard Fork

The Ezra Klein Show

Moonshots with Peter Diamandis

No Priors: Artificial Intelligence | Technology | Startups

Latent Space: The AI Engineer Podcast

BG2Pod with Brad Gerstner and Bill Gurley