8 citations · 23 across the 18 of their papers we have counts for
19 papers
On The Effectiveness-Fluency Trade-Off In LLM Conditioning: A Systematic Study
Iuri Macocco, Pau Rodríguez, Arno Blaas +3
Controlling the output of Large Language Models (LLMs) is a central challenge for their reliable deployment, yet a clear understanding of the involved trade-offs remains elusive. C…
Uncertainty Quantification for LLM Function-Calling
Zihuiwen Ye, Lukas Aichberger, Michael Kirchhof +5
Large Language Models (LLMs) are increasingly deployed to autonomously solve real-world tasks. A key ingredient for this is the LLM Function-Calling paradigm, a widely used approac…
The Design Space of Tri-Modal Masked Diffusion Models
Louis Bethune, Victor Turrisi, Bruno Kacper Mlodozeniec +21
Discrete diffusion models have emerged as strong alternatives to autoregressive language models, with recent work initializing and fine-tuning a base unimodal model for bimodal gen…
Sparse Autoencoders are Capable LLM Jailbreak Mitigators
Yannick Assogba, Jacopo Cortellazzi, Javier Abad +3
Jailbreak attacks remain a persistent threat to large language model safety. We propose Context-Conditioned Delta Steering (CC-Delta), an SAE-based defense that identifies jailbrea…
GenCtrl -- A Formal Controllability Toolkit for Generative Models
Emily Cheng, Carmen Amo Alonso, Federico Danieli +4
As generative models become ubiquitous, there is a critical need for fine-grained control over the generation process. Yet, while controlled generation methods from prompting to fi…
SelfReflect: Can LLMs Communicate Their Internal Answer Distribution?
Michael Kirchhof, Luca Füger, Adam Goliński +4
The common approach to communicate a large language model's (LLM) uncertainty is to add a percentage number or a hedging word to its response. But is this all we can do? Instead of…
Revisiting Uncertainty Quantification Evaluation in Language Models: Spurious Interactions with Response Length Bias Results
Andrea Santilli, Adam Golinski, Michael Kirchhof +5
Uncertainty Quantification (UQ) in Language Models (LMs) is key to improving their safety and reliability. Evaluations often use metrics like AUROC to assess how well UQ methods (e…
LinEAS: End-to-end Learning of Activation Steering with a Distributional Loss
Pau Rodriguez, Michal Klein, Eleonora Gualdoni +5
The growing use of generative models in daily life calls for efficient mechanisms to control their generation, to e.g., produce safe content or provide users with tools to explore…
Controlling Language and Diffusion Models by Transporting Activations
Pau Rodriguez, Arno Blaas, Michal Klein +4
The increasing capabilities of large generative models and their ever more widespread deployment have raised concerns about their reliability, safety, and potential misuse. To addr…
Considerations for Distribution Shift Robustness of Diagnostic Models in Healthcare
Arno Blaas, Adam Goliński, Andrew Miller +3
We consider robustness to distribution shifts in the context of diagnostic models in healthcare, where the prediction target , e.g., the presence of a disease, is causally upstr…
Interpreting CLIP: Insights on the Robustness to ImageNet Distribution Shifts
Jonathan Crabbé, Pau Rodríguez, Vaishaal Shankar +2
What distinguishes robust models from non-robust ones? While for ImageNet distribution shifts it has been shown that such differences in robustness can be traced back predominantly…
The Role of Entropy and Reconstruction in Multi-View Self-Supervised Learning
Borja Rodríguez-Gálvez, Arno Blaas, Pau Rodríguez +5
The mechanisms behind the success of multi-view self-supervised learning (MVSSL) are not yet fully understood. Contrastive MVSSL methods have been studied through the lens of InfoN…
DUET: 2D Structured and Approximately Equivariant Representations
Xavier Suau, Federico Danieli, T. Anderson Keller +5
Multiview Self-Supervised Learning (MSSL) is based on learning invariances with respect to a set of input transformations. However, invariance partially or totally removes transfor…
Challenges of Adversarial Image Augmentations
Arno Blaas, Xavier Suau, Jason Ramapuram +2
Image augmentations applied during training are crucial for the generalization performance of image classifiers. Therefore, a large body of research has focused on finding the opti…
Adversarial Attacks on Graph Classification via Bayesian Optimisation
Xingchen Wan, Henry Kenlay, Binxin Ru +3
Graph neural networks, a popular class of models effective in a wide range of graph-based learning tasks, have been shown to be vulnerable to adversarial attacks. While the majorit…
On Invariance Penalties for Risk Minimization
Kia Khezeli, Arno Blaas, Frank Soboczenski +2
The Invariant Risk Minimization (IRM) principle was first proposed by Arjovsky et al. [2019] to address the domain generalization problem by leveraging data heterogeneity from diff…
Adversarial Robustness Guarantees for Gaussian Processes
Andrea Patane, Arno Blaas, Luca Laurenti +3
Gaussian processes (GPs) enable principled computation of model uncertainty, making them attractive for safety-critical applications. Such scenarios demand that GP decisions are no…
The Effect of Prior Lipschitz Continuity on the Adversarial Robustness of Bayesian Neural Networks
Arno Blaas, Stephen J. Roberts
It is desirable, and often a necessity, for machine learning models to be robust against adversarial attacks. This is particularly true for Bayesian models, as they are well-suited…
Adversarial Robustness Guarantees for Classification with Gaussian Processes
Arno Blaas, Andrea Patane, Luca Laurenti +3
We investigate adversarial robustness of Gaussian Process Classification (GPC) models. Given a compact subset of the input space enclosing a test point $x…