3 papers
cs.CL2026
Agreement in Representation Space for Open-Ended Self-Consistency
Paula Ontalvilla, Gorka Azkune, Aitor Ormazabal
Self-consistency improves LLM reasoning by sampling multiple outputs and selecting the most consistent answer, but existing formulations largely rely on exact matching and therefor…
cs.CL2026
Multimodal LLMs Do Not Compose Skills Optimally Across Modalities
Paula Ontalvilla, Aitor Ormazabal, Gorka Azkune
Skill composition is the ability to combine previously learned skills to solve new tasks. As neural networks acquire increasingly complex skills during their pretraining, it is not…
cs.CL2024
Improving the Efficiency of Visually Augmented Language Models
Paula Ontalvilla, Aitor Ormazabal, Gorka Azkune
Despite the impressive performance of autoregressive Language Models (LM) it has been shown that due to reporting bias, LMs lack visual knowledge, i.e. they do not know much about…