Showing 2026Show all
3 papers · 1 filter
cs.CL2026
Likelihood Ranking doesn't Scale Like Prompting in LLMs
Alessandro Bondielli, Lucia Passaro, Davide Bacciu +1
LLM evaluation is commonly performed either by prompting models to produce answers or by scoring candidate outputs with likelihood-based metrics. In multiple-choice QA, however, st…
cs.CL2026
Parts-of-Speech as Emergent Categories in SAE Latent Space
Alessandro Bondielli, Lucia Passaro, Serena Auriemma +1
Sparse AutoEncoders (SAEs) offer a promising way to inspect language model representations, but it is still unclear what kind of linguistic structure their latents expose. We use p…
cs.CL2026
From Vision to Language: Investigating Causal Information Flow in Multimodal Decision-Making
Davide Testa, Hugh Mee Wong, Alessandro Lenci +2
Vision-Language Models are commonly evaluated through their final predictions, but understanding whether these decisions are grounded in visual evidence requires tracing how visual…