Matthew Bozoukov

According to our database, Matthew Bozoukov authored at least 5 papers between 2025 and 2026.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of four.

2026

Are LLM Evaluators Really Narcissists? Sanity Checking Self-Preference Evaluations.

Dani Roytburg, Matthew Bozoukov, Matthew Nguyen, Mackenzie Puig-Hall, Narmeen Oozeer

CoRR, January, 2026

2025

Minimal and Mechanistic Conditions for Behavioral Self-Awareness in LLMs.

Matthew Bozoukov, Matthew Nguyen, Shubkarman Singh, Bart Bussmann, Patrick Leask

CoRR, November, 2025

Breaking the Mirror: Activation-Based Mitigation of Self-Preference in LLM Evaluators.

Dani Roytburg, Matthew Bozoukov, Matthew Nguyen, Jou Barzdukas, Simon Fu, Narmeen Oozeer

CoRR, September, 2025

Generating and Predicting Output Perturbations in Image Segmenters.

Matthew Bozoukov, Nguyen Anh Vu Doan, Bryan Donyanavard

Proceedings of the Design, Automation & Test in Europe Conference, 2025

Uncovering Branch-specialization in InceptionV1 using k sparse autoencoders.

Matthew Bozoukov

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2025