26 октября 2025 г.

ОБ ЭТОМ ЭПИЗОДЕ

This episode of "Two Minds, One Model" explores the critical concept of interpretability in AI systems, focusing on Anthropic's research paper "Toy Models of Superposition." Hosts John Jezl and Jon Rocha from Sonoma State University's Computer Science Department delve into why neural networks are often "black boxes" and what this means for AI safety and deployment.

Credits

Cover Art by Brianna Williams

TMOM Intro Music by Danny Meza

A special thank you to these talented artists for their contributions to the show.

—---------------------------------------------------

Links and Reference

Academic Papers

“Toy Models of Superposition” - Anthropic (December 2022)
"Alignment Faking in Large Language Models" - Anthropic (December 2024)
"Agentic Misalignment: How LLMs Could Be Insider Threats" - Anthropic (January 2025)

News

Harvard Business School study on companion chatbots

Misc

“Words are but vague shadows of the volumes we mean”' - Theodore Dreiser
3Blue1Brown video about vectors - https://www.youtube.com/shorts/FJtFZwbvkI4
GPT-3 parameter count Correction: https://en.wikipedia.org/wiki/GPT-3#:~:text=GPT%2D3%20has%20175%20billion,each%20parameter%20occupies%202%20bytes.
ImageNet: ImageNet: A Large-Scale Hierarchical Image Database

We mention Waymo a lot in this episode and felt it was important to link to their safety page: https://waymo.com/safety/

Abandoned Episode Titles

"404: Interpretation Not Found"

"Neurons Gone Wild: Spring Break Edition"

"These Aren't the Features You're Looking For”

"Bigger on the Inside"

Технология

английский

США

ИспользуйтеAPI, чтобы загрузить этот эпизод вJSON.

В ЭТОМ ЭПИЗОДЕ

# theodore dreiser

danny meza

brianna williams

# jon rocha

john jezl

Изменить эти теги

РАСШИФРОВКА 🔗

Are you the producer of this podcast? Add a podcast transcript

Need Audio-to-Text? Transcribe with Listen411 in Just 60 Seconds

ПОИСК ПРОШЛЫХ ЭПИЗОДОВ

Искать прошлые эпизоды Two Minds, One Model.

Вставьте эту панель поиска на свой сайт

ДРУГИЕ ЭПИЗОДЫ В ЭТОМ ПОДКАсте

Decomposing Superposition: Sparse Autoencoders for Neural Network Interpretability

4 ноября 2025 г.

This episode explores how sparse autoencoders can decode the phenomenon of superposition in neural networks, demonstrating that the seemingly impenetrable compression of features into neurons can be partially reversed to extract interpretable, causal features. The discussion centers on an Anthropic…

00:53:20

СОХРАНИТЬ

Клип

A Brief History of Time

6 октября 2025 г.

This premiere episode provides a comprehensive history of artificial intelligence development from the 1950s through the present day, tracing the cycles of excitement and disappointment ("summers and winters") that led to today's breakthrough moment with large language models. The hosts e…

01:11:09

СОХРАНИТЬ

Клип

34 Million Features Later: What Researchers Found Inside Claude's World Model

8 ноября 2025 г.

This episode explores how Anthropic researchers successfully scaled sparse autoencoders from toy models to Claude 3 Sonnet's 8 billion neurons, extracting 34 million interpretable features including ones for deception, sycophancy, and the famous Golden Gate Bridge example. The discussion emphasizes…

00:59:58

СОХРАНИТЬ

Клип

What if We Succeed?

7 октября 2025 г.

This episode explores why AI systems might develop harmful or deceptive behaviors even without malicious intent, examining concepts like convergent instrumental goals, alignment faking, and mesa optimization to explain how models pursuing benign objectives can still take problematic actions. The ho…

01:12:48

СОХРАНИТЬ

Клип

Отказ от ответственности: Подкаст и иллюстрации, встроенные в эту страницу, взяты из John Jezl and Jon Rocha, который является собственностью его владельца и не связан или не одобрен Listen Notes, Inc.