Picture for Moa Johansson

Moa Johansson

Chalmers University of Technology

Fact Recall, Heuristics or Pure Guesswork? Precise Interpretations of Language Models for Fact Completion

Add code
Oct 18, 2024
Viaarxiv icon

ACE: Abstractions for Communicating Efficiently

Add code
Sep 30, 2024
Viaarxiv icon

Setting the AI Agenda -- Evidence from Sweden in the ChatGPT Era

Add code
Sep 25, 2024
Viaarxiv icon

Learning Efficient Recursive Numeral Systems via Reinforcement Learning

Add code
Sep 11, 2024
Viaarxiv icon

Specify What? Enhancing Neural Specification Synthesis by Symbolic Methods

Add code
Jun 21, 2024
Figure 1 for Specify What? Enhancing Neural Specification Synthesis by Symbolic Methods
Figure 2 for Specify What? Enhancing Neural Specification Synthesis by Symbolic Methods
Figure 3 for Specify What? Enhancing Neural Specification Synthesis by Symbolic Methods
Figure 4 for Specify What? Enhancing Neural Specification Synthesis by Symbolic Methods
Viaarxiv icon

Can Large Language Models (or Humans) Distill Text?

Add code
Mar 25, 2024
Viaarxiv icon

Reasoning in Transformers -- Mitigating Spurious Correlations and Reasoning Shortcuts

Add code
Mar 17, 2024
Viaarxiv icon

The Effect of Scaling, Retrieval Augmentation and Form on the Factual Consistency of Language Models

Add code
Nov 02, 2023
Viaarxiv icon

Towards Learning Abstractions via Reinforcement Learning

Add code
Dec 28, 2022
Viaarxiv icon

Conjectures, Tests and Proofs: An Overview of Theory Exploration

Add code
Sep 07, 2021
Figure 1 for Conjectures, Tests and Proofs: An Overview of Theory Exploration
Figure 2 for Conjectures, Tests and Proofs: An Overview of Theory Exploration
Figure 3 for Conjectures, Tests and Proofs: An Overview of Theory Exploration
Figure 4 for Conjectures, Tests and Proofs: An Overview of Theory Exploration
Viaarxiv icon