Aly M. Kassem

14 papers A* 2C 1Journal 7Unranked 4
YearRankTypeTitle / Venue / Authors
2026 J jnl
Delta-Crosscoder: Robust Crosscoder Model Diffing in Narrow Fine-Tuning Regimes.
CoRR
Aly M. Kassem, Thomas Jiralerspong, Negar Rostamzadeh, Golnoosh Farnadi
2026 — conf
How Robust Are Router-LLMs? Analysis of the Fragility of LLM Routing Capabilities.
EACL (Volume 1: Long Papers)
Aly M. Kassem, Bernhard Schölkopf, Zhijing Jin
2025 — conf
ALPACA AGAINST VICUNA: Using LLMs to Uncover Memorization of LLMs.
NAACL (Long Papers)
Aly M. Kassem, Omar Mahmoud, Niloofar Mireshghallah, Hyunwoo Kim, Yulia Tsvetkov, Yejin Choi, Sherif Saad, Santu Rana
2025 J jnl
How Robust Are Router-LLMs? Analysis of the Fragility of LLM Routing Capabilities.
CoRR
Aly M. Kassem, Bernhard Schölkopf, Zhijing Jin
2025 A* conf
REVIVING YOUR MNEME: Predicting The Side Effects of LLM Unlearning and Fine-Tuning via Sparse Model Diffing.
EMNLP
Aly M. Kassem, Zhuan Shi, Negar Rostamzadeh, Golnoosh Farnadi
2025 J jnl
Reviving Your MNEME: Predicting The Side Effects of LLM Unlearning and Fine-Tuning via Sparse Model Diffing.
CoRR
Aly M. Kassem, Zhuan Shi, Negar Rostamzadeh, Golnoosh Farnadi
2024 J jnl
Alpaca against Vicuna: Using LLMs to Uncover Memorization of LLMs.
CoRR
Aly M. Kassem, Omar Mahmoud, Niloofar Mireshghallah, Hyunwoo Kim, Yulia Tsvetkov, Yejin Choi, Sherif Saad, Santu Rana
2024 — conf
Finding a Needle in the Adversarial Haystack: A Targeted Paraphrasing Approach For Uncovering Edge Cases with Minimal Distribution Distortion.
EACL (1)
Aly M. Kassem, Sherif Saad
2024 J jnl
Finding a Needle in the Adversarial Haystack: A Targeted Paraphrasing Approach For Uncovering Edge Cases with Minimal Distribution Distortion.
CoRR
Aly M. Kassem, Sherif Saad
2023 J jnl
An ensemble transformer-based model for Arabic sentiment analysis.
Soc. Netw. Anal. Min.
Omar Mohamed, Aly M. Kassem, Ali Ashraf, Salma Jamal, Ensaf Hussein Mohamed
2023 J jnl
Mitigating Approximate Memorization in Language Models via Dissimilarity Learned Policy.
CoRR
Aly M. Kassem
2023 A* conf
Preserving Privacy Through Dememorization: An Unlearning Technique For Mitigating Memorization Risks In Language Models.
EMNLP
Aly M. Kassem, Omar Mahmoud, Sherif Saad
2022 C conf
In the Identification of Arabic Dialects: A Loss Function Ensemble Learning Based-Approach.
MEDI
Salma Jamal, Salma Khaled, Aly M. Kassem, Ayaalla Eltabey, Alaa Osama, Samah Mohamed, Mustafa A. Elattar
2022 — conf
On The Arabic Dialects' Identification: Overcoming Challenges of Geographical Similarities Between Arabic dialects and Imbalanced Datasets.
WANLP@EMNLP
Salma Jamal, Aly M. Kassem, Omar Mohamed, Ali Ashraf