11 papers · 1 filter
Activation-Aware Weight Tensorization: A Calibration-Time Preconditioner for Tensor-Network LLM Compression
Alessandro Beatini, Marco Maronese, Emanuele Rodolà
Post-training tensor-network compression replaces Transformer linear layers with Tensor Train (TT) or Tree Tensor Network (TTN) operators, but standard decompositions minimize weig…
MoGround: Measuring and Mitigating Modality Distraction in Vision-Language Models
Luca Zhou, Bo Zhao, Rose Yu +2
We release MoGround, a vision-language dataset spanning four visual domains in which the answer to every question is guaranteed to be available from exactly one modality. This guar…
Weight Pair Encoding: Inducing a Smaller Grammar in Neural Network Weights
Irene Tallini, Daniele Solombrino, Alberto Cazzaniga +1
We show that neural network weights can be explicilty fintuned to admit a smaller grammar. Weight Pair Encoding (WeightPE) does so by placing a lossy Re-Pair compressor inside a st…
On Emergent Capabilities and Model Merging
Luca Zhou, Emanuele Rodolà
Fine-tuned checkpoints and adapters now fill public repositories, and the most common operation applied to these artifacts is model merging: arithmetic on their weights that assemb…
The Undetected Damage of Quantization on Retrieval and How to Fix It
Luca Zhou, Alessandro Zirilli, Daniele Solombrino +2
We show that a quantized model that keeps its classification accuracy still changes to of its top-1 retrieval results, and that aggregate ranking metrics reveal only pa…
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…
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…
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…
Communicating Sound Through Natural Language
Emanuele Rossi, Emanuele Rodolà
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…
Not All Latent Spaces Are Flat: Hyperbolic Concept Control
Maria Rosaria Briglia, Simone Facchiano, Paolo Cursi +6
As modern text-to-image (T2I) models draw closer to synthesizing highly realistic content, the threat of unsafe content generation grows, and it becomes paramount to exercise contr…
Cluster-driven Graph Federated Learning over Multiple Domains
Debora Caldarola, Massimiliano Mancini, Fabio Galasso +3
Federated Learning (FL) deals with learning a central model (i.e. the server) in privacy-constrained scenarios, where data are stored on multiple devices (i.e. the clients). The ce…