Study notebook

Press N to jot a timestamped note or B to pin a moment while watching.

On the Emergence of Representational Geometry from Natural Language Statistics

Yasaman BahriHarvard University

Research Scientist

1:08:42·2026·research level·Research Seminar

Yasaman Bahri of Google DeepMind discusses the spontaneous emergence of intricate geometric structures in large language models, reflecting the semantic and conceptual organization of natural language.

AI & ML Engineering Language Models & NLP Complex Systems & Scaling Algorithms & Data Structures large language models

New Technologies in Mathematics Seminar 11/4/2026

Source: Harvard CMSA via YouTubeWatch on YouTube

LecturerAll lecturers →

Research Scientist

Comments

Add a comment...Sign in

Related lectures

Marin: An Open Lab For Frontier AI

Percy Liang · Stanford University

Shared subfield · Language Models & NLP & Complex Systems & Scaling

Revisiting Main Memory-Based Timing Attacks in the Context of PIM (DSN'25 Talk)

Nisa Bostanci · ETH Zürich

Shared subfield · Algorithms & Data Structures & Complex Systems & Scaling

When Prediction Becomes Policy: What Do Risk Models Actually Learn?

Shion Guha · UC Berkeley

Shared subfield · Language Models & NLP & Algorithms & Data Structures

Atomic Facts to Structured Knowledge: Rethinking Unlearning & Jailbreaking in Large Language Models

Rongzhe Wei · Google Research & TechTalks

Shared subfield · Language Models & NLP & large language models

Training Sand to Think: Artificial General Intelligence & Future of Physics

Adam Brown · Perimeter Institute

Shared subfield · Language Models & NLP & large language models

Enhanced and Efficient Reasoning in Large Language Models

Leslie Valiant · UC Berkeley

Shared subfield · Language Models & NLP & large language models