Julius Broomfield

Undergrad student, Georgia Institute of Technology

Joined

May 2024

Names

Julius Broomfield (Preferred)

Emails

****@gatech.edu (Confirmed)

Personal Links

Google Scholar

DBLP

Semantic Scholar

Career & Education History

Undergrad student

Georgia Institute of Technology (gatech.edu)

2022 – Present

Advisors, Relations & Conflicts

No relations added

Expertise

Multimodal Models

Present

Large Language Models

Present

Computational Social Science

Present

Robustness

Present

Fairness

Present

Publications

From Where Do Language Models See? Surfacing and Mitigating Blind Spots via Latent Need Discovery
Julius Broomfield, Kartik Sharma
- Eval Eval @ ACL 2026 Poster
- Readers: Everyone
A Thousand Words or An Image: Studying the Influence of Persona Modality in Multimodal LLMs
Julius Broomfield, Kartik Sharma, Srijan Kumar
- ICML 2025 R2-FM Workshop Poster
- Readers: Everyone
A Thousand Words or An Image: Studying the Influence of Persona Modality in Multimodal LLMs
Julius Broomfield, Kartik Sharma, Srijan Kumar
- CoRR 2025
- Readers: Everyone
The Structural Safety Generalization Problem
Tom Gibbs, Julius Broomfield, George Ingebretsen, Ethan Kosak-Hine, Tia Nasir, Jason Zhang, Reihaneh Iranmanesh, Sara Pieri, Reihaneh Rabbany, Kellin Pelrine
- SafeGenAi Poster
- Readers: Everyone
Decompose, Recompose, and Conquer: Multi-modal LLMs are Vulnerable to Compositional Adversarial Attacks in Multi-Image Queries
Julius Broomfield, George Ingebretsen, Reihaneh Iranmanesh, Sara Pieri, Ethan Kosak-Hine, Tom Gibbs, Reihaneh Rabbany, Kellin Pelrine
- NeurIPS 2024 Workshop RBFM Poster
- Readers: Everyone
Decompose, Recompose, and Conquer: Multi-modal LLMs are Vulnerable to Compositional Adversarial Attacks in Multi-Image Queries
Julius Broomfield, George Ingebretsen, Reihaneh Iranmanesh, Sara Pieri, Ethan Kosak-Hine, Tom Gibbs, Reihaneh Rabbany, Kellin Pelrine
- Red Teaming GenAI Workshop @ NeurIPS'24 Poster
- Readers: Everyone
Emerging Vulnerabilities in Frontier Models: Multi-Turn Jailbreak Attacks
Tom Gibbs, Ethan Kosak-Hine, George Ingebretsen, Jason Zhang, Julius Broomfield, Sara Pieri, Reihaneh Iranmanesh, Reihaneh Rabbany, Kellin Pelrine
- CoRR 2024
- Readers: Everyone

Julius Broomfield

Names

Emails

Personal Links

Career & Education History

Advisors, Relations & Conflicts

Expertise

Publications

From Where Do Language Models See? Surfacing and Mitigating Blind Spots via Latent Need Discovery

A Thousand Words or An Image: Studying the Influence of Persona Modality in Multimodal LLMs

A Thousand Words or An Image: Studying the Influence of Persona Modality in Multimodal LLMs

The Structural Safety Generalization Problem

Decompose, Recompose, and Conquer: Multi-modal LLMs are Vulnerable to Compositional Adversarial Attacks in Multi-Image Queries

Decompose, Recompose, and Conquer: Multi-modal LLMs are Vulnerable to Compositional Adversarial Attacks in Multi-Image Queries

Emerging Vulnerabilities in Frontier Models: Multi-Turn Jailbreak Attacks

Co-Authors