23 citations · 30 across the 20 of their papers we have counts for
20 papers
Subliminal Clocks: Latent Time Modelling in Diffusion Language Models
Maximo Eduardo Rulli, Thomas Vaitses Fontanari, Simone Petruzzi +9
Diffusion Language Models (DLMs) have recently emerged as a promising alternative to autoregressive models. Unlike standard diffusion-based approaches, DLMs are not explicitly cond…
PHALAR: Phasors for Learned Musical Audio Representations
Davide Marincione, Michele Mancusi, Giorgio Strano +4
Stem retrieval, the task of matching missing stems to a given audio submix, is a key challenge currently limited by models that discard temporal information. We introduce PHALAR, a…
Multi-Way Representation Alignment
Akshit Achara, Tatiana Gaintseva, Mateo Mahaut +5
The Platonic Representation Hypothesis suggests that independently trained neural networks converge to increasingly similar latent spaces. However, current strategies for mapping t…
Membership and Dataset Inference Attacks on Large Audio Generative Models
Jakub Proboszcz, Paweł Kochanski, Karol Korszun +5
Generative audio models, based on diffusion and autoregressive architectures, have advanced rapidly in both quality and expressiveness. This progress, however, raises pressing copy…
Language Models are Injective and Hence Invertible
Giorgos Nikolaou, Tommaso Mencattini, Donato Crisostomi +3
Transformer components such as non-linear activations and normalization are inherently non-injective, suggesting that different inputs could map to the same output and prevent exac…
Two-Scale Latent Dynamics for Recurrent-Depth Transformers
Francesco Pappone, Donato Crisostomi, Emanuele Rodolà
Recurrent-depth transformers scale test-time compute by iterating latent computations before emitting tokens. We study the geometry of these iterates and argue for a simple, two-sc…
On Task Vectors and Gradients
Luca Zhou, Daniele Solombrino, Donato Crisostomi +4
Task arithmetic has emerged as a simple yet powerful technique for model merging, enabling the combination of multiple finetuned models into one. Despite its empirical success, a c…
Implicit Inversion turns CLIP into a Decoder
Antonio D'Orazio, Maria Rosaria Briglia, Donato Crisostomi +3
CLIP is a discriminative model trained to align images and text in a shared embedding space. Due to its multimodal structure, it serves as the backbone of many generative pipelines…
Update Your Transformer to the Latest Release: Re-Basin of Task Vectors
Filippo Rinaldi, Giacomo Capitani, Lorenzo Bonicelli +6
Foundation models serve as the backbone for numerous specialized models developed through fine-tuning. However, when the underlying pretrained model is updated or retrained (e.g.,…
Mergenetic: a Simple Evolutionary Model Merging Library
Adrian Robert Minut, Tommaso Mencattini, Andrea Santilli +2
Model merging allows combining the capabilities of existing models into a new one - post hoc, without additional training. This has made it increasingly popular thanks to its low c…
Activation Patching for Interpretable Steering in Music Generation
Simone Facchiano, Giorgio Strano, Donato Crisostomi +4
Understanding how large audio models represent music, and using that understanding to steer generation, is both challenging and underexplored. Inspired by mechanistic interpretabil…
LoopGen: Training-Free Loopable Music Generation
Davide Marincione, Giorgio Strano, Donato Crisostomi +2
Loops--short audio segments designed for seamless repetition--are central to many music genres, particularly those rooted in dance and electronic styles. However, current generativ…
MERGE: Efficient Evolutionary Merging on Consumer-grade GPUs
Tommaso Mencattini, Adrian Robert Minut, Donato Crisostomi +2
Evolutionary model merging enables the creation of high-performing multi-task models but remains computationally prohibitive for consumer hardware. We introduce MERGE, an effic…
Humanity's Last Exam
Long Phan, Alice Gatti, Ziwen Han +1168
Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achi…
Task Singular Vectors: Reducing Task Interference in Model Merging
Antonio Andrea Gargiulo, Donato Crisostomi, Maria Sofia Bucarelli +3
Task Arithmetic has emerged as a simple yet effective method to merge models without additional training. However, by treating entire networks as flat parameter vectors, it overloo…
ATM: Improving Model Merging by Alternating Tuning and Merging
Luca Zhou, Daniele Solombrino, Donato Crisostomi +3
Model merging has emerged as a cost-efficient approximation to multitask learning. Among merging strategies, task arithmetic is notable for its simplicity and effectiveness. In thi…
: Cycle-Consistent Multi-Model Merging
Donato Crisostomi, Marco Fumero, Daniele Baieri +2
In this paper, we present a novel data-free method for merging neural networks in weight space. Differently from most existing works, our method optimizes for the permutations of n…
From Charts to Atlas: Merging Latent Spaces into One
Donato Crisostomi, Irene Cannistraci, Luca Moschella +4
Models trained on semantically related datasets and tasks exhibit comparable inter-sample relations within their latent spaces. We investigate in this study the aggregation of such…
Efficient Generation of Multimodal Fluid Simulation Data
Daniele Baieri, Donato Crisostomi, Stefano Esposito +2
In this work, we introduce an efficient generation procedure to produce synthetic multi-modal datasets of fluid simulations. The procedure can reproduce the dynamics of fluid flows…
Metric Based Few-Shot Graph Classification
Donato Crisostomi, Simone Antonelli, Valentino Maiorca +3
Many modern deep-learning techniques do not work without enormous datasets. At the same time, several fields demand methods working in scarcity of data. This problem is even more c…