Aly M. Kassem @_AKassem

Exploration over Exploitation. RA @Mila_Quebec, Research Fellow @UniofOxford. MSc @UWindsor. Interested in Adversarial attacks, security & reliability of LLMs Joined April 2014

Tweets

132
Followers

92
Following

900
Likes

6K

Aly M. Kassem @_AKassem

a month ago

🎉 Thrilled that our work has been accepted at #EMNLP2025 (Main Conference)! TL;DR: We propose a framework to predict & explain unintended side effects in models (e.g., emergent toxicity, forgotten knowledge) using OOD data. Huge thanks to @gfarnadi, @negar_rz, and Zhuan Shi 🚀

Aly M. Kassem @_AKassem

2 months ago

1 1 7 2K 4

Download Image

0 2 18 2K 4

Aly M. Kassem @_AKassem

2 months ago

We observed similar biased behavior when evaluating LLM routers, even with commercial solutions such as Amazon Bedrock By biases, I mean that the router tends to favor certain categories or keywords, consistently directing them to the more powerful model arxiv.org/abs/2504.07113

Jeremy Howard @jeremyphoward

2 months ago

69 67 920 92K 258

0 0 3 211 0

Wenhu Chen @WenhuChen

4 months ago

Finally, the crazy weeks of NeurIPS ddl, ICCV rebuttal and EMNLP ddl have passed. Now it's time to take a rest till Sep!

3 2 102 11K 6

Aly M. Kassem @_AKassem

6 months ago

Very useful thread — unfortunately, I learned it the hard way.

Max Forbes @maxforbes

6 months ago

Very useful thread — unfortunately, I learned it the hard way.

1 4 24 3K 28

0 0 0 198 1

Zhijing Jin@HLF2025 @ZhijingJin

6 months ago

Check out my mentee's latest work on LLM Router attack! First work on this topic to the best of our knowledge. Read our paper at: zhijing-jin.com/files/papers/2… Great job to @_AKassem!🎉 @MPI_IS @UofTCompSci