Bernard Ghanem

Towards Assessing and Characterizing the Semantic Robustness of Face Recognition

2022 · arXiv (Cornell University)

Deep Neural Networks (DNNs) lack robustness against imperceptible perturbations to their input. Face Recognition Models (FRMs) based on DNNs inherit this vulnerability. We propose a methodology for assessing and characterizing the robustness of FRMs against …

Robust Optimization as Data Augmentation for Large-scale Graphs

2020 · arXiv (Cornell University)

Data augmentation helps neural networks generalize better by enlarging the training set, but it remains an open question how to effectively augment graph data to enhance the performance of GNNs (Graph Neural Networks). While most …

Computationally Budgeted Continual Learning: What Does Matter?

2023 · arXiv (Cornell University)

Continual Learning (CL) aims to sequentially train models on streams of incoming data that vary in distribution by preserving previous knowledge while adapting to new data. Current CL literature focuses on restricted access to previously …

Online Distillation with Continual Learning for Cyclic Domain Shifts

2023

In recent years, online distillation has emerged as a powerful technique for adapting real-time deep neural networks on the fly using a slow, but accurate teacher model. However, a major challenge in online distillation is …

Don’t FREAK Out: A Frequency-Inspired Approach to Detecting Backdoor Poisoned Samples in DNNs

2023

In this paper we investigate the frequency sensitivity of Deep Neural Networks (DNNs) when presented with clean samples versus poisoned samples. Our analysis shows significant disparities in frequency sensitivity between these two types of samples. …

From Categories to Classifiers: Name-Only Continual Learning by Exploring the Web

2023 · arXiv (Cornell University)

Continual Learning (CL) often relies on the availability of extensive annotated datasets, an assumption that is unrealistically time-consuming and costly in practice. We explore a novel paradigm termed name-only continual learning where time and cost …

An Embarrassingly Simple Defense Against LLM Abliteration Attacks

2025 · arXiv (Cornell University)

Large language models (LLMs) are typically aligned to refuse harmful instructions through safety fine-tuning. A recent attack, termed abliteration, identifies and suppresses the single latent direction most responsible for refusal behavior, thereby enabling models to …

AVA: Attentive VLM Agent for Mastering StarCraft II

2026

We introduce AVACraft, a multimodal StarCraft II benchmark supporting both Multi-Agent Reinforcement Learning (MARL) and Vision-Language Model (VLM) paradigms. Unlike SMAC-family environments that rely on abstract state representations and exclude VLMs, AVACraft provides RGB visuals, …

Beyond the Last Answer: Your Reasoning Trace Uncovers More than You Think

2025 · arXiv (Cornell University)

Large Language Models (LLMs) leverage step-by-step reasoning to solve complex problems. Standard evaluation practice involves generating a complete reasoning trace and assessing the correctness of the final answer presented at its conclusion. In this paper, …