Posts by Collection

portfolio

publications

CEA-List@EvalLLM2024: prompting a large language model or fine-tuning a smaller one?

Published in EvalLLM2024 , 2024

The EvalLLM2024 challenge aims to evaluate the results of few-shot approaches to information extraction in French. Our contribution to this challenge tests two approaches: one exploits the available annotated data in the prompt of an LLM (in context learning) while the other fine-tunes a generic entity recognition model (GLiNER) by exploiting the annotated data. Our experiments show that this second approach obtains the best results, especially when enriched by a data augmentation step exploiting the annotation guide and LLMs for the generation of synthetic examples.

Download Paper

GLiDRE : Generalist model for Document-level Relation Extraction

Published in EGC - Atelier TextMine, 2025

This work adapts the GLiNER model for document-level relation extraction in French, as part of the TextMine 2025 challenge. Enhancements include pretraining on a subset of OSCAR dataset, local representations inspired by ATLOP, and optimized prediction thresholds. Results demonstrate modest performance but the model demonstrates potential in low-resource scenarios.

Download Paper

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.