Picture for Zhixue Zhao

Zhixue Zhao

Label Set Optimization via Activation Distribution Kurtosis for Zero-shot Classification with Generative Models

Add code
Oct 24, 2024
Viaarxiv icon

Rulebreakers Challenge: Revealing a Blind Spot in Large Language Models' Reasoning with Formal Logic

Add code
Oct 21, 2024
Figure 1 for Rulebreakers Challenge: Revealing a Blind Spot in Large Language Models' Reasoning with Formal Logic
Figure 2 for Rulebreakers Challenge: Revealing a Blind Spot in Large Language Models' Reasoning with Formal Logic
Figure 3 for Rulebreakers Challenge: Revealing a Blind Spot in Large Language Models' Reasoning with Formal Logic
Figure 4 for Rulebreakers Challenge: Revealing a Blind Spot in Large Language Models' Reasoning with Formal Logic
Viaarxiv icon

Can We Reverse In-Context Knowledge Edits?

Add code
Oct 16, 2024
Figure 1 for Can We Reverse In-Context Knowledge Edits?
Figure 2 for Can We Reverse In-Context Knowledge Edits?
Figure 3 for Can We Reverse In-Context Knowledge Edits?
Figure 4 for Can We Reverse In-Context Knowledge Edits?
Viaarxiv icon

Language-specific Calibration for Pruning Multilingual Language Models

Add code
Aug 26, 2024
Viaarxiv icon

Detecting Edited Knowledge in Language Models

Add code
May 04, 2024
Viaarxiv icon

Comparing Explanation Faithfulness between Multilingual and Monolingual Fine-tuned Language Models

Add code
Mar 19, 2024
Viaarxiv icon

ReAGent: A Model-agnostic Feature Attribution Method for Generative Language Models

Add code
Feb 07, 2024
Viaarxiv icon

Lighter, yet More Faithful: Investigating Hallucinations in Pruned Large Language Models for Abstractive Summarization

Add code
Nov 15, 2023
Viaarxiv icon

Incorporating Attribution Importance for Improving Faithfulness Metrics

Add code
May 17, 2023
Viaarxiv icon

On the Impact of Temporal Concept Drift on Model Explanations

Add code
Oct 17, 2022
Figure 1 for On the Impact of Temporal Concept Drift on Model Explanations
Figure 2 for On the Impact of Temporal Concept Drift on Model Explanations
Figure 3 for On the Impact of Temporal Concept Drift on Model Explanations
Figure 4 for On the Impact of Temporal Concept Drift on Model Explanations
Viaarxiv icon