Publications

2026

1. Under Review

   Reward-Gated On-Policy Distillation

   Sadegh Akhondzadeh, Vijay Lingam, Atula Tejaswi, and 3 more authors
2. Under Review

   Should We Pay This Much for Robustness? Efficient Proxy Certificates with Marginal Guarantees

   Soroush H Zargarbashi^*, Sadegh Akhondzadeh^*, and Aleksandar Bojchevski
3. Under Review

   CATS: Conformalized Adaptive Test-time Scaling

   Sadegh Akhondzadeh, Soroush H Zargarbashi, Simone Antonelli, and 1 more author

   Oral @ Catch, Adapt, and Operate (CAO) ICLR 2026 Workshop

   Best Poster

   Best Poster Award at the Catch, Adapt, and Operate (CAO) ICLR 2026 Workshop.

4. Under Review

   How Test-Time Training Undermines Existing Safety Guardrails

   Simone Antonelli^*, Sadegh Akhondzadeh^*, and Aleksandar Bojchevski

   Oral @ Principled Design for Trustworthy AI ICLR 2026 Workshop
5. ICML

   Front-Loaded Robust Conformal Prediction: Heavy Calibration, Minimal Test-Time Cost

   Soroush H Zargarbashi, Sadegh Akhondzadeh, and Aleksandar Bojchevski

   In International Conference on Machine Learning, ICML, 2026
6. ICLR

   EvA: Evolutionary Attacks on Graphs

   Sadegh Akhondzadeh, Soroush H Zargarbashi, Jimin Cao, and 1 more author

   In International Conference on Learning Representations, ICLR, 2026

2025

1. EMNLP

   KurTail: Kurtosis-based LLM Quantization

   Sadegh Akhondzadeh, Aleksandar Bojchevski, Evangelos Eleftheriou, and 1 more author

   In Conference on Empirical Methods in Natural Language Processing, EMNLP. Also at the ICLR 2025 SLLM Workshop , 2025
2. NeurIPS

   One Sample is Enough to Make Conformal Prediction Robust

   Soroush H Zargarbashi, Sadegh Akhondzadeh, and Aleksandar Bojchevski

   In Advances in Neural Information Processing Systems, NeurIPS, 2025

2024

1. ICLR

   Rethinking Label Poisoning for GNNs: Pitfalls and Attacks

   Vijay Lingam, Sadegh Akhondzadeh, and Aleksandar Bojchevski

   In The Twelfth International Conference on Learning Representations, ICLR, 2024

   PDF
2. ICML

   Robust Yet Efficient Conformal Prediction Sets

   Soroush H Zargarbashi, Sadegh Akhondzadeh, and Aleksandar Bojchevski

   In International Conference on Machine Learning, ICML, 2024

   PDF

2023

1. AISTATS

   Probing Graph Representation

   Sadegh Akhondzadeh, Vijay Lingam, and Aleksandar Bojchevski

   In International Conference on Artificial Intelligence and Statistics AISTATS, 2023

   Code Website

2021

1. WEBSCI 2021

   Wide-AdGraph: Detecting Ad Trackers with a Wide Dependency Chain Graph

   Amir Hossein Kargaran, Sadegh Akhondzadeh, Mohammad Reza Heidarpour, and 3 more authors

   In 13th ACM Web Science Conference, 2021

   Abs arXiv PDF Video Code

   Websites use third-party ads and tracking services to deliver targeted ads and collect information about users that visit them. These services put users’ privacy at risk, and that is why users’ demand for blocking these services is growing. Most of the blocking solutions rely on crowd-sourced filter lists manually maintained by a large community of users. In this work, we seek to simplify the update of these filter lists by combining different websites through a large-scale graph connecting all resource requests made over a large set of sites. The features of this graph are extracted and used to train a machine learning algorithm with the aim of detecting ads and tracking resources. As our approach combines different information sources, it is more robust toward evasion techniques that use obfuscation or changing the usage patterns. We evaluate our work over the Alexa top-10K websites and find its accuracy to be 96.1% biased and 90.9% unbiased with high precision and recall. It can also block new ads and tracking services, which would necessitate being blocked by further crowd-sourced existing filter lists. Moreover, the approach followed in this paper sheds light on the ecosystem of third-party tracking and advertising.

2020

1. Frontiers 2020

   Automated detection of autism spectrum disorder using a convolutional neural network

   Zeinab Sherkatghanad^*, Sadegh Akhondzadeh^*, Soorena Salari, and 5 more authors

   Frontiers in Neuroscience, 2020

   HTML PDF
2. Evaluating Sparse Interpretable Word Embeddings for Biomedical Domain

   Mohammad Amin Samadi, Sadegh Akhondzadeh, Sayed Jalal Zahabi, and 3 more authors

   arXiv preprint arXiv:2005.05114, 2020

   arXiv
3. ICML-Compbio

   Learning Cancer Progression Network from Mutation Allele Frequencies

   Sadegh Akhondzadeh, Alireza Omidi, Zeinab Maleki, and 3 more authors

   In ICML workshop on Computational Biology, 2020

   PDF