Manifold Steering Llm Control Via Geometry

May 15, 2026
Media Summary: In this AI Research Roundup episode, Alex discusses the paper: ' Support BrainOmega ☕ Buy Me a Coffee: Stripe: ... Modify the behavior or the personality of a model at inference time, without fine-tuning or prompt engineering. Read the blog post ...

Manifold Steering Llm Control Via Geometry - Detailed Analysis & Overview

In this AI Research Roundup episode, Alex discusses the paper: ' Support BrainOmega ☕ Buy Me a Coffee: Stripe: ... Modify the behavior or the personality of a model at inference time, without fine-tuning or prompt engineering. Read the blog post ... In this AI Research Roundup episode, Alex discusses the paper: 'Do Sparse Autoencoders Capture Concept State-of-the-art foundation models are often seen as black boxes: we send a prompt in and we get out our - often useful - answer. My name is Artem, I'm a neuroscience PhD student at Harvard University. Website and Social links:

In this AI Research Roundup episode, Alex discusses the paper: 'Why When we talk about Deep Learning, we usually talk about layers, weights, and data. But underneath all that is the Space. Part of my research at Cal Poly, San Luis Obispo. The prompt used was: Prompt 3 (Alternating color stack w correction "1. Clay Reid, Allen Institute for Brain Science Targeted Discovery in Brain Data. This video summarizes the research by Eric Bigelow, Daniel Wurgaft, and colleagues from Goodfire AI, Harvard, NTT Research, ...

Photo Gallery

Manifold Steering: LLM Control via Geometry
Neural Manifolds: Generative AI Geometry Tutorial
Steering LLM Behavior Without Fine-Tuning
How SAEs Map Concept Manifolds in LLMs
Steering vectors: tailor LLMs without training. Part I: Theory (Interpretability Series)
Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior (May 202
Steering vectors: tailor LLMs without training. Part II: Code (Interpretability Series)
Neural manifolds - The Geometry of Behaviour
What is a manifold?
Yonghyeon Lee - A geometric take on motion manifold learning from demonstration
SPLIT: A Unified Framework for LLM Steering
Manifold Learning. The Geometric Constraints of Mixed-Curvature Product Manifolds. Geometric DL.

Related Celebrities

Manifold Steering: LLM Control via Geometry Neural Manifolds: Generative AI Geometry Tutorial Steering LLM Behavior Without Fine-Tuning How SAEs Map Concept Manifolds in LLMs Steering vectors: tailor LLMs without training. Part I: Theory (Interpretability Series) Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior (May 202 Steering vectors: tailor LLMs without training. Part II: Code (Interpretability Series) Neural manifolds - The Geometry of Behaviour Rekomendasi Erek 62 Untuk Pemain Baru 3 Alasan Mengapa Erek 2 Selalu Menjadi Pilihan Pemain Judi Cara Menghasilkan Togel Orang Buta: Menguak Rahasia Menghitung Angka Menguntungkan Jika Kamu Mau Menang, Baca Angka Mimpi 2D Melahirkan Mengenal Rahasia Kecoak Togel Yang Sebelumnya Tak Terungkap Rahasia Tafsir Mimpi 2D 10001 Untuk Menang Judi Online Buku Mimpi Beras 2D: Prediksi Tepat Angka Jitu Dengan Tepat Mengenal Alasan Orang Mengalami Mimpi Suami Selingkuh Yang Paling Unik Dan Membuat Keprihatinan
Sponsored
Sponsored
View Detailed Profile
Manifold Steering: LLM Control via Geometry

Manifold Steering: LLM Control via Geometry

In this AI Research Roundup episode, Alex discusses the paper: '

Neural Manifolds: Generative AI Geometry Tutorial

Neural Manifolds: Generative AI Geometry Tutorial

Support BrainOmega ☕ Buy Me a Coffee: https://buymeacoffee.com/brainomega Stripe: ...

Sponsored
Steering LLM Behavior Without Fine-Tuning

Steering LLM Behavior Without Fine-Tuning

Modify the behavior or the personality of a model at inference time, without fine-tuning or prompt engineering. Read the blog post ...

How SAEs Map Concept Manifolds in LLMs

How SAEs Map Concept Manifolds in LLMs

In this AI Research Roundup episode, Alex discusses the paper: 'Do Sparse Autoencoders Capture Concept

Steering vectors: tailor LLMs without training. Part I: Theory (Interpretability Series)

Steering vectors: tailor LLMs without training. Part I: Theory (Interpretability Series)

State-of-the-art foundation models are often seen as black boxes: we send a prompt in and we get out our - often useful - answer.

Sponsored
Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior (May 202

Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior (May 202

Title:

Steering vectors: tailor LLMs without training. Part II: Code (Interpretability Series)

Steering vectors: tailor LLMs without training. Part II: Code (Interpretability Series)

See Part I for an intro into

Neural manifolds - The Geometry of Behaviour

Neural manifolds - The Geometry of Behaviour

My name is Artem, I'm a neuroscience PhD student at Harvard University. Website and Social links: https://kirsanov.ai/ ...

What is a manifold?

What is a manifold?

A visual explanation and definition of

Yonghyeon Lee - A geometric take on motion manifold learning from demonstration

Yonghyeon Lee - A geometric take on motion manifold learning from demonstration

A

SPLIT: A Unified Framework for LLM Steering

SPLIT: A Unified Framework for LLM Steering

In this AI Research Roundup episode, Alex discusses the paper: 'Why

Manifold Learning. The Geometric Constraints of Mixed-Curvature Product Manifolds. Geometric DL.

Manifold Learning. The Geometric Constraints of Mixed-Curvature Product Manifolds. Geometric DL.

When we talk about Deep Learning, we usually talk about layers, weights, and data. But underneath all that is the Space.

LLM-Guided Robotic Manipulation

LLM-Guided Robotic Manipulation

Part of my research at Cal Poly, San Luis Obispo. The prompt used was: Prompt 3 (Alternating color stack w correction "1.

Computational Geometry of Cortical Networks: Manifold-Embedding Analysis of Multi-Site Recordings...

Computational Geometry of Cortical Networks: Manifold-Embedding Analysis of Multi-Site Recordings...

Clay Reid, Allen Institute for Brain Science https://simons.berkeley.edu/talks/clay-reid-3-20-18 Targeted Discovery in Brain Data.

How Belief Dynamics Control LLMs: ICL and Activation Steering Unified

How Belief Dynamics Control LLMs: ICL and Activation Steering Unified

This video summarizes the research by Eric Bigelow, Daniel Wurgaft, and colleagues from Goodfire AI, Harvard, NTT Research, ...

Mapping LLM Reasoning via Geometric Trajectories

Mapping LLM Reasoning via Geometric Trajectories

In this AI Research Roundup episode, Alex discusses the paper: '