Copertina

Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Feucht, Sheridan, Haklay, Tal, Bhalla, Usha, Wurgaft, Daniel, Rager, Can, Sarfati, Raphaël, Merullo, Jack, McGrath, Thomas, Lewis, Owen, Lubana, Ekdeep Singh, Fel, Thomas, Geiger, Atticus
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Artificial Intelligence Computation and Language 68T50 I.2.7
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

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

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

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

The Shape of Beliefs: Geometry, Dynamics, and Interventions along Representation Manifolds of Language Models' Posteriors
di: Sarfati, Raphaël, et al.
Pubblicazione: (2026)

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

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

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

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

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

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

Evaluating Sparse Autoencoders: From Shallow Design to Matching Pursuit
di: Costa, Valérie, et al.
Pubblicazione: (2025)

Vector Arithmetic in Concept and Token Subspaces
di: Feucht, Sheridan, et al.
Pubblicazione: (2025)

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

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

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

Priors in Time: Missing Inductive Biases for Language Model Interpretability
di: Lubana, Ekdeep Singh, et al.
Pubblicazione: (2025)

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

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

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

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

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

Comparisons of the 1995 and 1998 coral bleaching events on the patch reefs of San Salvador Island, Bahamas
di: Thomas A. McGrath
Pubblicazione: (2003)

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

Archetypal SAE: Adaptive and Stable Dictionary Learning for Concept Extraction in Large Vision Models
di: Fel, Thomas, et al.
Pubblicazione: (2025)

Sparks of Explainability: Recent Advancements in Explaining Large Vision Models
di: Fel, Thomas
Pubblicazione: (2025)

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

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

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

Language Models Implement Simple Word2Vec-style Vector Arithmetic
di: Merullo, Jack, et al.
Pubblicazione: (2023)

Into the Rabbit Hull: From Task-Relevant Concepts in DINO to Minkowski Geometry
di: Fel, Thomas, et al.
Pubblicazione: (2025)

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

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

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

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

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

Bucketing the Good Apples: A Method for Diagnosing and Improving Causal Abstraction
di: Puyin, Li, et al.
Pubblicazione: (2026)

Discriminative Feature Attributions: Bridging Post Hoc Explainability and Inherent Interpretability
di: Bhalla, Usha, et al.
Pubblicazione: (2023)

The Dual-Route Model of Induction
di: Feucht, Sheridan, et al.
Pubblicazione: (2025)

Erasing Conceptual Knowledge from Language Models
di: Gandikota, Rohit, et al.
Pubblicazione: (2024)

Token Erasure as a Footprint of Implicit Vocabulary Items in LLMs
di: Feucht, Sheridan, et al.
Pubblicazione: (2024)