collaborators

8 papers

cs.LG2026

Hard or Just Unreached? Diagnosing the Sampling Blind Spot in Math-Reasoning Difficulty Estimation

Luca Zhou, Sajel Shah, Emanuele Rodolà +1

Math and science reasoning benchmarks rely on pass@k, the fraction of sampled chains that reach gold, as the canonical per-example difficulty signal. The same signal drives RL with…

cs.LG2026

Steering Vectors are an Adversarial Attack Surface

Abzal Aidakhmetov, Donato Crisostomi, Tommaso Mencattini +3

Activation steering has become a popular way to control Large Language Model (LLM) behavior without fine-tuning. Since the technique is plug-and-play, users share datasets and prec…

cs.LG2026

How Neural Losses Shape VAE Latents

Giorgio Strano, Luca Cerovaz, Michele Mancusi +2

Modern VAEs are rarely trained with the pointwise likelihood implied by the standard -VAE objective. In practice, pointwise reconstruction is often combined with perceptual and…

cs.LG2026

Communicating Sound Through Natural Language

Emanuele Rossi, Emanuele RodolÃ

Natural language is widely used to describe, prompt, and control audio systems, but rarely serves as the representation carrying audio itself. We introduce lexical acoustic coding…

cs.CV2026

Zero-Shot Quantization via Weight-Space Arithmetic

Daniele Solombrino, Antonio Andrea Gargiulo, Alessandro Zirilli +3

We show that robustness to post-training quantization (PTQ) is a transferable direction in weight space. We call this direction the quantization vector: extracted from a donor task…

cs.CL2026

Multi-objective Evolutionary Merging Enables Efficient Reasoning Models

Mario Iacobelli, Adrian Robert Minut, Tommaso Mencattini +5

Reasoning models achieve strong performance on complex problems by leveraging long chains of thought, but this deliberate reasoning incurs substantial inference-time cost. The Long…