Imagen de Portada

Mechanistically analyzing the effects of fine-tuning on procedurally defined tasks

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Jain, Samyak, Kirk, Robert, Lubana, Ekdeep Singh, Dick, Robert P., Tanaka, Hidenori, Grefenstette, Edward, Rocktäschel, Tim, Krueger, David Scott
Formato:	Preprint
Publicado:	2023
Materias:	Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

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

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

What Makes and Breaks Safety Fine-tuning? A Mechanistic Study
por: Jain, Samyak, et al.
Publicado: (2024)

Compositional Capabilities of Autoregressive Transformers: A Study on Synthetic, Interpretable Tasks
por: Ramesh, Rahul, et al.
Publicado: (2023)

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

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

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

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

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

Towards an Understanding of Stepwise Inference in Transformers: A Synthetic Graph Navigation Model
por: Khona, Mikail, et al.
Publicado: (2024)

minimax: Efficient Baselines for Autocurricula in JAX
por: Jiang, Minqi, et al.
Publicado: (2023)

Emergence of Hidden Capabilities: Exploring Learning Dynamics in Concept Space
por: Park, Core Francisco, et al.
Publicado: (2024)

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

Swing-by Dynamics in Concept Learning and Compositional Generalization
por: Yang, Yongyi, et al.
Publicado: (2024)

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

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

Investigating Non-Transitivity in LLM-as-a-Judge
por: Xu, Yi, et al.
Publicado: (2025)

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

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

Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models
por: Ruis, Laura, et al.
Publicado: (2024)

ICLR: In-Context Learning of Representations
por: Park, Core Francisco, et al.
Publicado: (2024)

Detecting High-Stakes Interactions with Activation Probes
por: McKenzie, Alex, et al.
Publicado: (2025)

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

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

Infusion: Shaping Model Behavior by Editing Training Data via Influence Functions
por: Rosser, J, et al.
Publicado: (2026)

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

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

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

Scaling Opponent Shaping to High Dimensional Games
por: Khan, Akbir, et al.
Publicado: (2023)

The Impact of Off-Policy Training Data on Probe Generalisation
por: Kirch, Nathalie, et al.
Publicado: (2025)

From Isolation to Entanglement: When Do Interpretability Methods Identify and Disentangle Known Concepts?
por: Mueller, Aaron, et al.
Publicado: (2025)

Reward Model Ensembles Help Mitigate Overoptimization
por: Coste, Thomas, et al.
Publicado: (2023)

Force Dipole Interactions in Tubular Fluid Membranes
por: Jain, Samyak, et al.
Publicado: (2023)

Nuclear stability and the Fold Catastrophe
por: Jain, Samyak, et al.
Publicado: (2023)

Tunneling half-lives in macroscopic-microscopic picture
por: Jain, Samyak, et al.
Publicado: (2024)

Catastrophe theoretic approach to the Higgs Mechanism
por: Jain, Samyak, et al.
Publicado: (2023)

Understanding the Effects of RLHF on LLM Generalisation and Diversity
por: Kirk, Robert, et al.
Publicado: (2023)

Interaction Dynamics as a Reward Signal for LLMs
por: Gooding, Sian, et al.
Publicado: (2025)

Learning When to Plan: Efficiently Allocating Test-Time Compute for LLM Agents
por: Paglieri, Davide, et al.
Publicado: (2025)

Debating with More Persuasive LLMs Leads to More Truthful Answers
por: Khan, Akbir, et al.
Publicado: (2024)