Publications

LLM-Safety Evaluations Lack Robustness
Tim Beyer
Sophie Xhonneux
Simon Geisler
Gauthier Gidel
Leo Schwinn
Stephan Günnemann
In this paper, we argue that current safety alignment research efforts for large language models are hindered by many intertwined sources of… (voir plus) noise, such as small datasets, methodological inconsistencies, and unreliable evaluation setups. This can, at times, make it impossible to evaluate and compare attacks and defenses fairly, thereby slowing progress. We systematically analyze the LLM safety evaluation pipeline, covering dataset curation, optimization strategies for automated red-teaming, response generation, and response evaluation using LLM judges. At each stage, we identify key issues and highlight their practical impact. We also propose a set of guidelines for reducing noise and bias in evaluations of future attack and defense papers. Lastly, we offer an opposing perspective, highlighting practical reasons for existing limitations. We believe that addressing the outlined problems in future research will improve the field's ability to generate easily comparable results and make measurable progress.
2025-03-01
arXiv (publié)
doi.org
arxiv.org
Negotiative Alignment: Embracing Disagreement to Achieve Fairer Outcomes -- Insights from Urban Studies
Rashid A. Mushkani
Hugo Berard
Shin (Alexandre) Koseki
2025-03-01
arXiv (publié)
doi.org
arxiv.org
Normalizing Spinal Cord Compression Measures in Degenerative Cervical Myelopathy.
Sandrine Bédard
Jan Valosek
Maryam Seif
Armin Curt
Simon Schading-Sassenhausen
Nikolai Pfender
P. Freund
Markus Hupp
Julien Cohen-Adad
2025-03-01
The spine journal (publié)
doi.org
PRISM: High-Resolution & Precise Counterfactual Medical Image Generation using Language-guided Stable Diffusion
Amar Kumar
Anita Kriz
Mohammad Havaei
Tal Arbel
Developing reliable and generalizable deep learning systems for medical imaging faces significant obstacles due to spurious correlations, da… (voir plus)ta imbalances, and limited text annotations in datasets. Addressing these challenges requires architectures robust to the unique complexities posed by medical imaging data. The rapid advancements in vision-language foundation models within the natural image domain prompt the question of how they can be adapted for medical imaging tasks. In this work, we present PRISM, a framework that leverages foundation models to generate high-resolution, language-guided medical image counterfactuals using Stable Diffusion. Our approach demonstrates unprecedented precision in selectively modifying spurious correlations (the medical devices) and disease features, enabling the removal and addition of specific attributes while preserving other image characteristics. Through extensive evaluation, we show how PRISM advances counterfactual generation and enables the development of more robust downstream classifiers for clinically deployable solutions. To facilitate broader adoption and research, we make our code publicly available at https://github.com/Amarkr1/PRISM.
2025-03-01
arXiv (publié)
doi.org
RL4Med-DDPO: Reinforcement Learning for Controlled Guidance Towards Diverse Medical Image Generation using Vision-Language Foundation Models
Parham Saremi
Amar Kumar
Mohammed Mohammed
Zahra Tehrani Nasab
Tal Arbel
2025-03-01
arXiv (publié)
doi.org
arxiv.org
Self-adaptive cyber defense for sustainable IoT: A DRL-based IDS optimizing security and energy efficiency
Saeid Jamshidi
Ashkan Amirnia
Amin Nikanjam
Kawser Wazed Nafi
Foutse Khomh
Samira Keivanpour
2025-03-01
Journal of Network and Computer Applications (publié)
doi.org
SemEval-2025 Task 11: Bridging the Gap in Text-Based Emotion Detection
Shamsuddeen Hassan Muhammad
Nedjma OUSIDHOUM
Idris Abdulmumin
Seid Muhie Yimam
Jan Philip Wahle
Terry Lima Ruas
Meriem Beloucif
Christine de Kock
Tadesse Belay
Ibrahim Ahmad
Nirmal Surange
Daniela Teodorescu
David Ifeoluwa Adelani
Alham Fikri Aji
Felermino Ali
Vladimir Araujo
Abinew Ayele
Oana Ignat
Alexander Panchenko
Yi Zhou … (voir 1 de plus)
Saif M. Mohammad
2025-03-01
arXiv (publié)
doi.org
arxiv.org
A three-state coupled Markov switching model for COVID-19 outbreaks across Quebec based on hospital admissions
Dirk Douwes-Schultz
Alexandra M. Schmidt
Yannan Shen
David Buckeridge
2025-03-01
The Annals of Applied Statistics (publié)
doi.org
arxiv.org
Tractable Representations for Convergent Approximation of Distributional HJB Equations
Julie Alhosh
Harley Wiltzer
David Meger
2025-03-01
arXiv (publié)
doi.org
arxiv.org
Unveiling Inefficiencies in LLM-Generated Code: Toward a Comprehensive Taxonomy
Altaf Allah Abbassi
Leuson Da Silva
Amin Nikanjam
Foutse Khomh
2025-03-01
arXiv (publié)
doi.org
arxiv.org
PRISM: High-Resolution&Precise Counterfactual Medical Image Generation using Language-guided Stable Diffusion
Amar Kumar
Anita Kriz
Mohammad Havaei
Tal Arbel
Developing reliable and generalizable deep learning systems for medical imaging faces significant obstacles due to spurious correlations, da… (voir plus)ta imbalances, and limited text annotations in datasets. Addressing these challenges requires architectures robust to the unique complexities posed by medical imaging data. The rapid advancements in vision-language foundation models within the natural image domain prompt the question of how they can be adapted for medical imaging tasks. In this work, we present PRISM, a framework that leverages foundation models to generate high-resolution, language-guided medical image counterfactuals using Stable Diffusion. Our approach demonstrates unprecedented precision in selectively modifying spurious correlations (the medical devices) and disease features, enabling the removal and addition of specific attributes while preserving other image characteristics. Through extensive evaluation, we show how PRISM advances counterfactual generation and enables the development of more robust downstream classifiers for clinically deployable solutions. To facilitate broader adoption and research, we make our code publicly available at https://github.com/Amarkr1/PRISM.
2025-02-28
ArXiv (prépublication)
arxiv.org
arxiv.org
Steering Large Language Model Activations in Sparse Spaces
Reza Bayat
Ali Rahimi-Kalahroudi
Mohammad Pezeshki
Sarath Chandar
Pascal Vincent
A key challenge in AI alignment is guiding large language models (LLMs) to follow desired behaviors at test time. Activation steering, which… (voir plus) modifies internal model activations during inference, offers a potential solution. However, prior work in dense activation spaces struggles with superposition, wherein multiple features become entangled, limiting interpretability and precise control. In contrast, sparse representations provide an untapped opportunity for more interpretable behavior modulation. In this work, we introduce sparse activation steering (SAS), a method that leverages sparse autoencoders (SAEs) to steer LLM behavior in sparse spaces. By isolating behavior-specific features through a contrastive prompt-pairing approach, we define a set of features that can selectively reinforce or suppress behaviors. Experiments on Gemma 2 LLMs show that SAS vectors enable nuanced behavioral modulation and finer-grained control. Furthermore, scaling SAEs improves monosemanticity of SAS vectors, suggesting more reliable and interpretable interventions.
2025-02-28
ArXiv (prépublication)
doi.org
arxiv.org