Cover Image

The Shape of Beliefs: Geometry, Dynamics, and Interventions along Representation Manifolds of Language Models' Posteriors

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Sarfati, Raphaël, Bigelow, Eric, Wurgaft, Daniel, Boppana, Siddharth, Merullo, Jack, Geiger, Atticus, Lewis, Owen, McGrath, Tom, Lubana, Ekdeep Singh
Format:	Preprint
Published:	2026
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space
by: Bigelow, Eric, et al.
Published: (2026)

Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought
by: Boppana, Siddharth, et al.
Published: (2026)

Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior
by: Wurgaft, Daniel, et al.
Published: (2026)

Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts
by: Feucht, Sheridan, et al.
Published: (2026)

Do Sparse Autoencoders Capture Concept Manifolds?
by: Bhalla, Usha, et al.
Published: (2026)

Are language models aware of the road not taken? Token-level uncertainty and hidden state dynamics
by: Zur, Amir, et al.
Published: (2025)

Features as Rewards: Scalable Supervision for Open-Ended Tasks via Interpretability
by: Prasad, Aaditya Vikram, et al.
Published: (2026)

Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering
by: Bigelow, Eric, et al.
Published: (2025)

Towards Reliable Evaluation of Behavior Steering Interventions in LLMs
by: Pres, Itamar, et al.
Published: (2024)

From Memorization to Reasoning in the Spectrum of Loss Curvature
by: Merullo, Jack, et al.
Published: (2025)

In-Context Learning Strategies Emerge Rationally
by: Wurgaft, Daniel, et al.
Published: (2025)

How Do Transformers Learn Variable Binding in Symbolic Programs?
by: Wu, Yiwei, et al.
Published: (2025)

Competition Dynamics Shape Algorithmic Phases of In-Context Learning
by: Park, Core Francisco, et al.
Published: (2024)

In-Context Learning Dynamics with Random Binary Sequences
by: Bigelow, Eric J., et al.
Published: (2023)

Projecting Assumptions: The Duality Between Sparse Autoencoders and Concept Geometry
by: Hindupur, Sai Sumedh R., et al.
Published: (2025)

How Do LLMs Persuade? Linear Probes Can Uncover Persuasion Dynamics in Multi-Turn Conversations
by: Jaipersaud, Brandon, et al.
Published: (2025)

Abrupt Learning in Transformers: A Case Study on Matrix Completion
by: Gopalani, Pulkit, et al.
Published: (2024)

Axiomatic Causal Interventions for Reverse Engineering Relevance Computation in Neural Retrieval Models
by: Chen, Catherine, et al.
Published: (2024)

Firefly swarms: What models for what physics?
by: Sarfati, Raphael
Published: (2024)

Tracking and triangulating firefly flashes in field recordings
by: Sarfati, Raphael
Published: (2024)

Emergence of Hierarchical Emotion Organization in Large Language Models
by: Zhao, Bo, et al.
Published: (2025)

From Flat to Hierarchical: Extracting Sparse Representations with Matching Pursuit
by: Costa, Valérie, et al.
Published: (2025)

Binomial confidence intervals for rare events: importance of defining margin of error relative to magnitude of proportion
by: McGrath, Owen, et al.
Published: (2021)

A General Simulation-Based Optimisation Framework for Multipoint Constant-Stress Accelerated Life Tests
by: McGrath, Owen, et al.
Published: (2025)

A General Simulation‐Based Optimisation Framework for Multipoint Constant‐Stress Accelerated Life Tests
by: Owen McGrath, et al.
Published: (2025)

Evaluating Open-Source Sparse Autoencoders on Disentangling Factual Knowledge in GPT-2 Small
by: Chaudhary, Maheep, et al.
Published: (2024)

Supporting Data: Linking Geodetically Resolved Uplift to Long-Term Orogenic Exhumation in the Southern Alps, New Zealand
by: McGrath, Jack, et al.
Published: (2025)

Representation Shattering in Transformers: A Synthetic Study with Knowledge Editing
by: Nishi, Kento, et al.
Published: (2024)

Constructing Interpretable Features from Compositional Neuron Groups
by: Shafran, Or, et al.
Published: (2025)

Uncovering Conceptual Blindspots in Generative Image Models Using Sparse Autoencoders
by: Bohacek, Matyas, et al.
Published: (2025)

Analyzing (In)Abilities of SAEs via Formal Languages
by: Menon, Abhinav, et al.
Published: (2024)

I Have No Mouth, and I Must Rhyme: Uncovering Internal Phonetic Representations in LLaMA 3.2
by: McLaughlin, Oliver, et al.
Published: (2025)

Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention
by: Huang, Jing, et al.
Published: (2026)

Recurrent Neural Networks Learn to Store and Generate Sequences using Non-Linear Representations
by: Csordás, Róbert, et al.
Published: (2024)

A Percolation Model of Emergence: Analyzing Transformers Trained on a Formal Language
by: Lubana, Ekdeep Singh, et al.
Published: (2024)

Compositional Abilities Emerge Multiplicatively: Exploring Diffusion Models on a Synthetic Task
by: Okawa, Maya, et al.
Published: (2023)

How Causal Abstraction Underpins Computational Explanation
by: Geiger, Atticus, et al.
Published: (2025)

RAVEL: Evaluating Interpretability Methods on Disentangling Language Model Representations
by: Huang, Jing, et al.
Published: (2024)

From Directions to Regions: Decomposing Activations in Language Models via Local Geometry
by: Shafran, Or, et al.
Published: (2026)

On Linear Representations and Pretraining Data Frequency in Language Models
by: Merullo, Jack, et al.
Published: (2025)