Pasquale Minervini

Universal Properties of Activation Sparsity in Modern Large Language Models

Non-Euclidean Harmonic Losses

Fast and Expressive Multi-Token Prediction with Probabilistic Circuits

Learning GUI Grounding with Spatial Reasoning from Visual Feedback

OpenSIR: Open-Ended Self-Improving Reasoner

PiCSAR: Probabilistic Confidence Selection And Ranking for Reasoning Chains

With Great Backbones Comes Great Adversarial Transferability

Neurosymbolic Diffusion Models

Noiser: Bounded Input Perturbations for Attributing Large Language Models

Is Complex Query Answering Really Complex?

When Can Proxies Improve the Sample Complexity of Preference Learning?

FLARE: Faithful Logic-Aided Reasoning and Exploration

An Auditing Test to Detect Behavioral Shift in Language Models

DeCoRe: Decoding by Contrasting Retrieval Heads to Mitigate Hallucinations

TuBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning

Low-Rank Compression of Language Models Via Differentiable Rank Selection

Is Complex Query Answering Really Complex?