Výsledky vyhledávání - "Franzon, Francesca"

Report

Principles of semantic and functional efficiency in grammatical patterning

Grammatical features such as number and gender serve two central functions in human languages. While they encode salient semantic attributes like numerosity and animacy, they also offload sentence processing cost by predictably linking words together

Externí odkaz: http://arxiv.org/abs/2410.15865

Zobrazit plný text záznamu

Report

Unnatural language processing: How do language models handle machine-generated prompts?

Autor: Kervadec, Corentin, Franzon, Francesca, Baroni, Marco

Language model prompt optimization research has shown that semantically and grammatically well-formed manually crafted prompts are routinely outperformed by automatically generated token sequences with no apparent meaning or syntactic structure, incl

Externí odkaz: http://arxiv.org/abs/2310.15829

Zobrazit plný text záznamu

Report

Cross-Domain Image Captioning with Discriminative Finetuning

Autor: Dessì, Roberto, Bevilacqua, Michele, Gualdoni, Eleonora, Rakotonirina, Nathanael Carraz, Franzon, Francesca, Baroni, Marco

Neural captioners are typically trained to mimic human-generated references without optimizing for any specific communication goal, leading to problems such as the generation of vague captions. In this paper, we show that fine-tuning an out-of-the-bo

Externí odkaz: http://arxiv.org/abs/2304.01662

Zobrazit plný text záznamu

Report

Referential communication in heterogeneous communities of pre-trained visual deep networks

Autor: Mahaut, Matéo, Franzon, Francesca, Dessì, Roberto, Baroni, Marco

As large pre-trained image-processing neural networks are being embedded in autonomous agents such as self-driving cars or robots, the question arises of how such systems can communicate with each other about the surrounding world, despite their diff

Externí odkaz: http://arxiv.org/abs/2302.08913

Zobrazit plný text záznamu

Report

Communication breakdown: On the low mutual intelligibility between human and neural captioning

Autor: Dessì, Roberto, Gualdoni, Eleonora, Franzon, Francesca, Boleda, Gemma, Baroni, Marco

We compare the 0-shot performance of a neural caption-based image retriever when given as input either human-produced captions or captions generated by a neural captioner. We conduct this comparison on the recently introduced ImageCoDe data-set (Kroj

Externí odkaz: http://arxiv.org/abs/2210.11512

Zobrazit plný text záznamu