Assessing skeptical views of interpretability research

Автор: Chris Potts

Загружено: 2025-11-10

Просмотров: 4689

Описание:

Stanford AI Lab Faculty Lunch, November 7, 2025. Updated version of https://web.stanford.edu/~cgpotts/blo...

0:59 - Severance
1:45 - Explainable AI, Anthropic Interp, Stanford Interp
5:15 - Interpretability methods: Attribution, Probes, Interventions
15:27 - Skeptical positions
16:42 - "Interpretability cannot be achieved"
18:32 - "Interpretability is merely analysis"
21:14 - "Analysis is overrated"
27:33 - "Interpretability is not leading to improvements"
30:04 - "Interpretability is not helping with AI safety"
36:08 - Summary, and Aryaman's sweatshirt

Assessing skeptical views of interpretability research

Доступные форматы для скачивания:

Скачать видео mp4

Информация по загрузке:

Скачать аудио mp3

Похожие видео

Stanford AI Club: Jeff Dean on Important AI Trends

Stanford AI Club: Jeff Dean on Important AI Trends

LLM и GPT - как работают большие языковые модели? Визуальное введение в трансформеры

LLM и GPT - как работают большие языковые модели? Визуальное введение в трансформеры

Why are prompt optimizers still so underrated?

Why are prompt optimizers still so underrated?

Causal Mechanistic Interpretability (Stanford lecture 1) - Atticus Geiger

Causal Mechanistic Interpretability (Stanford lecture 1) - Atticus Geiger

Как внимание стало настолько эффективным [GQA/MLA/DSA]

Как внимание стало настолько эффективным [GQA/MLA/DSA]

Why Language Models Hallucinate - Adam Kalai

Why Language Models Hallucinate - Adam Kalai

"How to measure intelligence?" | Six researchers debate

The Problem with A.I. Slop! - Computerphile

The Problem with A.I. Slop! - Computerphile

He Co-Invented the Transformer. Now: Continuous Thought Machines [Llion Jones / Luke Darlow]

He Co-Invented the Transformer. Now: Continuous Thought Machines [Llion Jones / Luke Darlow]

Finding linguistic structure in large language models

Finding linguistic structure in large language models

AI, Machine Learning, Deep Learning and Generative AI Explained

AI, Machine Learning, Deep Learning and Generative AI Explained

Сооснователь OpenAI о Будущем и Настоящем в AI. Подкаст на Русском - Илья Суцкевер

Сооснователь OpenAI о Будущем и Настоящем в AI. Подкаст на Русском - Илья Суцкевер

Melanie Mitchell, Evaluating Cognitive Capacities in AI Systems | Natural Philosophy Symposium 2025

Melanie Mitchell, Evaluating Cognitive Capacities in AI Systems | Natural Philosophy Symposium 2025

ДНК создал Бог? Самые свежие научные данные о строении. Как работает информация для жизни организмов

ДНК создал Бог? Самые свежие научные данные о строении. Как работает информация для жизни организмов

The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li

The Godmother of AI on jobs, robots & why world models are next | Dr. Fei-Fei Li

Flow-Matching vs Diffusion Models explained side by side

Flow-Matching vs Diffusion Models explained side by side

Sleeper Agents in Large Language Models - Computerphile

Sleeper Agents in Large Language Models - Computerphile

The Physics of A.I.

The Physics of A.I.

Что ошибочно пишут в книгах об ИИ [Двойной спуск]

Что ошибочно пишут в книгах об ИИ [Двойной спуск]

Yann LeCun | Self-Supervised Learning, JEPA, World Models, and the future of AI

Yann LeCun | Self-Supervised Learning, JEPA, World Models, and the future of AI