A cargo del Dr. Santiago Acevedo
Seminario con día y horario especial a realizarse el jueves 28 de agosto a las 11h.
El mismo estará a cargo del Dr. Santiago Acevedo, Scuola internazionale superiore di studi avanzati (SISSA), Trieste.
Su charla se titula:
An approach to identify the most semantically informative deep representations of text and images
Se envía a continuación un breve resumen:
Deep neural networks are known to develop similar representations for semantically related data, even when they belong to different domains, such as an image and its description, or the same text in different languages. We present a method for quantitatively investigating this phenomenon by measuring the relative information content of the representations of semantically related data and probing how it is encoded into multiple tokens of large language models (LLMs) and vision transformers. Looking first at how LLMs process pairs of translated sentences, we identify inner “semantic” layers containing the most language-transferable information. We find moreover that, on these layers, a larger LLM (DeepSeek-V3) extracts significantly more general information than a smaller one (Llama3.1-8B). Semantic information is spread across many tokens and it is characterized by long-distance correlations between tokens and by a causal left-to-right (i.e., past-future) asymmetry. We also identify layers encoding semantic information within visual transformers. We show that caption representations in the semantic layers of LLMs predict visual representations of the corresponding images. We observe significant and model-dependent information asymmetries between image and text representations.
El evento tendrá lugar en el auditorio ¨Prof. Dr. Luis N. Epele¨ del IFLP, sito en la diagonal 113 entre 63 y 64.
Adjuntamos también flyer del evento del cual agradeceremos su difusión.
Esperamos contar con su presencia.