Image de couverture de livre

Talking Heads: Understanding Inter-layer Communication in Transformer Language Models

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Merullo, Jack, Eickhoff, Carsten, Pavlick, Ellie
Format:	Preprint
Publié:	2024
Sujets:	Computation and Language Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Circuit Component Reuse Across Tasks in Transformer Language Models
par: Merullo, Jack, et autres
Publié: (2023)

Language Models Implement Simple Word2Vec-style Vector Arithmetic
par: Merullo, Jack, et autres
Publié: (2023)

The Same But Different: Structural Similarities and Differences in Multilingual Language Modeling
par: Zhang, Ruochen, et autres
Publié: (2024)

How Do Language Models Compose Functions?
par: Khandelwal, Apoorv, et autres
Publié: (2025)

Transferring Linear Features Across Language Models With Model Stitching
par: Chen, Alan, et autres
Publié: (2025)

What is an "Abstract Reasoner"? Revisiting Experiments and Arguments about Large Language Models
par: Yun, Tian, et autres
Publié: (2025)

Paths Not Taken: Understanding and Mending the Multilingual Factual Recall Pipeline
par: Lu, Meng, et autres
Publié: (2025)

Does CLIP Bind Concepts? Probing Compositionality in Large Image Models
par: Lewis, Martha, et autres
Publié: (2022)

Does Training on Synthetic Data Make Models Less Robust?
par: Zhang, Lingze, et autres
Publié: (2025)

mOthello: When Do Cross-Lingual Representation Alignment and Cross-Lingual Transfer Emerge in Multilingual Models?
par: Hua, Tianze, et autres
Publié: (2024)

Transformer Mechanisms Mimic Frontostriatal Gating Operations When Trained on Human Working Memory Tasks
par: Traylor, Aaron, et autres
Publié: (2024)

How Do Vision-Language Models Process Conflicting Information Across Modalities?
par: Hua, Tianze, et autres
Publié: (2025)

On Linear Representations and Pretraining Data Frequency in Language Models
par: Merullo, Jack, et autres
Publié: (2025)

Dual Process Learning: Controlling Use of In-Context vs. In-Weights Strategies with Weight Forgetting
par: Anand, Suraj, et autres
Publié: (2024)

Is This Just Fantasy? Language Model Representations Reflect Human Judgments of Event Plausibility
par: Lepori, Michael A., et autres
Publié: (2025)

From Prediction to Understanding: Will AI Foundation Models Transform Brain Science?
par: Serre, Thomas, et autres
Publié: (2025)

Stable Anisotropic Regularization
par: Rudman, William, et autres
Publié: (2023)

Outlier Dimensions Encode Task-Specific Knowledge
par: Rudman, William, et autres
Publié: (2023)

A Survey on LLM-Assisted Clinical Trial Recruitment
par: Ghosh, Shrestha, et autres
Publié: (2025)

Enhancing Retrieval-Augmented Generation: A Study of Best Practices
par: Li, Siran, et autres
Publié: (2025)

The Persona Paradox: Medical Personas as Behavioral Priors in Clinical Language Models
par: Abdullahi, Tassallah, et autres
Publié: (2026)

When Silence Is Golden: Can LLMs Learn to Abstain in Temporal QA and Beyond?
par: Zhou, Xinyu, et autres
Publié: (2026)

Interpretable Emergent Language Using Inter-Agent Transformers
par: Bhardwaj, Mannan
Publié: (2025)

MobileVLM: A Vision-Language Model for Better Intra- and Inter-UI Understanding
par: Wu, Qinzhuo, et autres
Publié: (2024)

Forgotten Polygons: Multimodal Large Language Models are Shape-Blind
par: Rudman, William, et autres
Publié: (2025)

Talking Nonsense: Probing Large Language Models' Understanding of Adversarial Gibberish Inputs
par: Cherepanova, Valeriia, et autres
Publié: (2024)

Talking to Yourself: Defying Forgetting in Large Language Models
par: Sun, Yutao, et autres
Publié: (2026)

Helpful or Harmful? Exploring the Efficacy of Large Language Models for Online Grooming Prevention
par: Prosser, Ellie, et autres
Publié: (2024)

Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought
par: Boppana, Siddharth, et autres
Publié: (2026)

What We Talk About When We Talk About LMs: Implicit Paradigm Shifts and the Ship of Language Models
par: Zhu, Shengqi, et autres
Publié: (2024)

Source-Modality Monitoring in Vision-Language Models
par: Hua, Etha Tianze, et autres
Publié: (2026)

Emergent Introspective Awareness in Large Language Models
par: Lindsey, Jack
Publié: (2026)

Understanding Inter-Session Intentions via Complex Logical Reasoning
par: Bai, Jiaxin, et autres
Publié: (2023)

Language Models Change Facts Based on the Way You Talk
par: Kearney, Matthew, et autres
Publié: (2025)

A Graph Talks, But Who's Listening? Rethinking Evaluations for Graph-Language Models
par: Petkar, Soham, et autres
Publié: (2025)

$100K or 100 Days: Trade-offs when Pre-Training with Academic Resources
par: Khandelwal, Apoorv, et autres
Publié: (2024)

Enhancing Answer Reliability Through Inter-Model Consensus of Large Language Models
par: Amiri-Margavi, Alireza, et autres
Publié: (2024)

Latent Multi-Head Attention for Small Language Models
par: Mehta, Sushant, et autres
Publié: (2025)

Identifying Semantic Induction Heads to Understand In-Context Learning
par: Ren, Jie, et autres
Publié: (2024)

Uncovering Intermediate Variables in Transformers using Circuit Probing
par: Lepori, Michael A., et autres
Publié: (2023)