Picture for Jack Merullo

Jack Merullo

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

Add code
Jun 13, 2024
Viaarxiv icon

Dual Process Learning: Controlling Use of In-Context vs. In-Weights Strategies with Weight Forgetting

May 28, 2024
Viaarxiv icon

Axiomatic Causal Interventions for Reverse Engineering Relevance Computation in Neural Retrieval Models

Add code
May 03, 2024
Viaarxiv icon

Transformer Mechanisms Mimic Frontostriatal Gating Operations When Trained on Human Working Memory Tasks

Feb 13, 2024
Viaarxiv icon

Characterizing Mechanisms for Factual Recall in Language Models

Add code
Oct 24, 2023
Figure 1 for Characterizing Mechanisms for Factual Recall in Language Models
Figure 2 for Characterizing Mechanisms for Factual Recall in Language Models
Figure 3 for Characterizing Mechanisms for Factual Recall in Language Models
Figure 4 for Characterizing Mechanisms for Factual Recall in Language Models
Viaarxiv icon

Circuit Component Reuse Across Tasks in Transformer Language Models

Add code
Oct 12, 2023
Figure 1 for Circuit Component Reuse Across Tasks in Transformer Language Models
Figure 2 for Circuit Component Reuse Across Tasks in Transformer Language Models
Figure 3 for Circuit Component Reuse Across Tasks in Transformer Language Models
Figure 4 for Circuit Component Reuse Across Tasks in Transformer Language Models
Viaarxiv icon

Language Models Implement Simple Word2Vec-style Vector Arithmetic

Add code
May 25, 2023
Figure 1 for Language Models Implement Simple Word2Vec-style Vector Arithmetic
Figure 2 for Language Models Implement Simple Word2Vec-style Vector Arithmetic
Figure 3 for Language Models Implement Simple Word2Vec-style Vector Arithmetic
Figure 4 for Language Models Implement Simple Word2Vec-style Vector Arithmetic
Viaarxiv icon

Does CLIP Bind Concepts? Probing Compositionality in Large Image Models

Add code
Dec 20, 2022
Figure 1 for Does CLIP Bind Concepts? Probing Compositionality in Large Image Models
Figure 2 for Does CLIP Bind Concepts? Probing Compositionality in Large Image Models
Figure 3 for Does CLIP Bind Concepts? Probing Compositionality in Large Image Models
Figure 4 for Does CLIP Bind Concepts? Probing Compositionality in Large Image Models
Viaarxiv icon

ezCoref: Towards Unifying Annotation Guidelines for Coreference Resolution

Add code
Oct 13, 2022
Figure 1 for ezCoref: Towards Unifying Annotation Guidelines for Coreference Resolution
Figure 2 for ezCoref: Towards Unifying Annotation Guidelines for Coreference Resolution
Figure 3 for ezCoref: Towards Unifying Annotation Guidelines for Coreference Resolution
Figure 4 for ezCoref: Towards Unifying Annotation Guidelines for Coreference Resolution
Viaarxiv icon

Linearly Mapping from Image to Text Space

Add code
Sep 30, 2022
Figure 1 for Linearly Mapping from Image to Text Space
Figure 2 for Linearly Mapping from Image to Text Space
Figure 3 for Linearly Mapping from Image to Text Space
Figure 4 for Linearly Mapping from Image to Text Space
Viaarxiv icon