activity
20242026
collaborators
Showing cs.LGShow all

7 papers · 1 filter

cs.LG2026

Logit Distance Bounds Representational Similarity

Beatrix M. G. Nielsen, Emanuele Marconato, Luigi Gresele +2

For a broad family of discriminative models that includes autoregressive language models, identifiability results imply that if two models induce the same conditional distributions…

cs.LG2026

Relational Linear Properties in Language Models: An Empirical Investigation

Giovanni Valer, Luigi Gresele, Marco Bronzini +1

Linear properties are ubiquitous in the representations of language models; however, testing them experimentally remains a challenging task. This work focuses on relational lineari…

cs.LG2026

Concise and Logically Consistent Conformal Sets for Neuro-Symbolic Concept-Based Models

Samuele Bortolotti, Emanuele Marconato, Andrea Pugnana +2

Neuro-Symbolic Concept-based Models (NeSy-CBMs) are a family of architectures that integrate neural networks with symbolic reasoning for enhanced reliability in high-stakes applica…

cs.LG2026

Shortcuts and Identifiability in Concept-based Models from a Neuro-Symbolic Lens

Samuele Bortolotti, Emanuele Marconato, Paolo Morettin +2

Concept-based Models are neural networks that learn a concept extractor to map inputs to high-level concepts and an inference layer to translate these into predictions. Ensuring th…

cs.LG2025

When Does Closeness in Distribution Imply Representational Similarity? An Identifiability Perspective

Beatrix M. G. Nielsen, Emanuele Marconato, Andrea Dittadi +1

When and why representations learned by different deep neural networks are similar is an active research topic. We choose to address these questions from the perspective of identif…

cs.LG2025

If Concept Bottlenecks are the Question, are Foundation Models the Answer?

Nicola Debole, Pietro Barbiero, Francesco Giannini +3

Concept Bottleneck Models (CBMs) are neural networks designed to conjoin high performance with ante-hoc interpretability. CBMs work by first mapping inputs (e.g., images) to high-l…