7 papers
Emergent Specialization in Populations of Self-Supervised Collaborative Vision Experts Without a Shared Gate or Cross-Agent Gradients
Aram Davtyan, Pablo Acuaviva, Sebastian Stapf +1
Can a population of neural networks develop a useful division of labor without a shared gate or gradients between agents? We study a setting where each network has its own weights,…
CNOT-Distance is NP-complete under all-to-all connectivity
Antonio Acuaviva, Arturo Acuaviva, Pablo Acuaviva
Given and an integer , we ask whether can be implemented by at most CNOT gates on fixed labelled wires with all-to-all connectivity. We prov…
Mathematical Discovery in the Wild: AI-Guided Proofs in Banach Space Theory
Antonio Acuaviva, Pablo Acuaviva
We investigate the capacity of current language models to contribute to mathematical research. In Banach space theory, AI systems generated key ideas and proofs for five new result…
A Unital Banach Algebra Which Is Not a Calkin Algebra
Antonio Acuaviva, Pablo Acuaviva
Let and denote the ideals of approximable and compact operators on a Banach space , respectively. We construct a unital Banach algebra of d…
COMiT: Learning Structured Visual Tokens through Sequential Communication
Aram Davtyan, Yusuf Sahin, Yasaman Haghighi +4
Discrete image tokenizers provide a sequential interface for vision and multimodal models, but are typically optimized for reconstruction or compression and therefore tend to encod…
Rethinking Visual Intelligence: Insights from Video Pretraining
Pablo Acuaviva, Aram Davtyan, Mariam Hassan +4
Large language models (LLMs) have demonstrated that large-scale pretraining enables systems to adapt rapidly to new problems with little supervision in the language domain. This su…
From Generation to Generalization: Emergent Few-Shot Learning in Video Diffusion Models
Pablo Acuaviva, Aram Davtyan, Mariam Hassan +4
Video Diffusion Models (VDMs) have emerged as powerful generative tools, capable of synthesizing high-quality spatiotemporal content. Yet, their potential goes far beyond mere vide…