Picture for Soheil Feizi

Soheil Feizi

RePanda: Pandas-powered Tabular Verification and Reasoning

Add code
Mar 14, 2025
Viaarxiv icon

Seeing What's Not There: Spurious Correlation in Multimodal LLMs

Add code
Mar 11, 2025
Viaarxiv icon

A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models

Add code
Feb 22, 2025
Figure 1 for A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models
Figure 2 for A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models
Figure 3 for A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models
Figure 4 for A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models
Viaarxiv icon

Almost AI, Almost Human: The Challenge of Detecting AI-Polished Writing

Add code
Feb 21, 2025
Viaarxiv icon

On Mechanistic Circuits for Extractive Question-Answering

Add code
Feb 12, 2025
Viaarxiv icon

RESTOR: Knowledge Recovery through Machine Unlearning

Add code
Oct 31, 2024
Figure 1 for RESTOR: Knowledge Recovery through Machine Unlearning
Figure 2 for RESTOR: Knowledge Recovery through Machine Unlearning
Figure 3 for RESTOR: Knowledge Recovery through Machine Unlearning
Figure 4 for RESTOR: Knowledge Recovery through Machine Unlearning
Viaarxiv icon

Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models

Add code
Oct 17, 2024
Figure 1 for Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
Figure 2 for Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
Figure 3 for Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
Figure 4 for Unearthing Skill-Level Insights for Understanding Trade-Offs of Foundation Models
Viaarxiv icon

IntCoOp: Interpretability-Aware Vision-Language Prompt Tuning

Add code
Jun 19, 2024
Viaarxiv icon

Endor: Hardware-Friendly Sparse Format for Offloaded LLM Inference

Add code
Jun 17, 2024
Figure 1 for Endor: Hardware-Friendly Sparse Format for Offloaded LLM Inference
Figure 2 for Endor: Hardware-Friendly Sparse Format for Offloaded LLM Inference
Figure 3 for Endor: Hardware-Friendly Sparse Format for Offloaded LLM Inference
Figure 4 for Endor: Hardware-Friendly Sparse Format for Offloaded LLM Inference
Viaarxiv icon

Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models

Add code
Jun 12, 2024
Figure 1 for Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models
Figure 2 for Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models
Figure 3 for Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models
Figure 4 for Understanding and Mitigating Compositional Issues in Text-to-Image Generative Models
Viaarxiv icon