Papers by Pierre-Yves Oudeyer
Recursive Training Loops in LLMs: How training data properties modulate distribution shift in generated data? (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) are increasingly used in the creation of online content, creating feedback loops as future generations of models will be trained on this synthetic data. |
| Approach: | They propose to use large language models to create feedback loops as future models are trained on this data. |
| Outcome: | The proposed model collapse effects are found to be detrimental to the results of recursive training on human datasets. |
Reinforcement Learning for Aligning Large Language Models Agents with Interactive Environments: Quantifying and Mitigating Prompt Overfitting (2025.findings-naacl)
Copied to clipboard
Mohamed Salim Aissi, Clément Romac, Thomas Carta, Sylvain Lamprier, Pierre-Yves Oudeyer, Olivier Sigaud, Laure Soulier, Nicolas Thome
| Challenge: | Reinforcement learning (RL) is a promising approach for aligning large language models knowledge with sequential decision-making tasks. |
| Approach: | They propose to use a contrastive loss framework to analyze the sensitivity of LLMs to prompt formulations following RL training in a textual environment. |
| Outcome: | The proposed framework improves the model's robustness and generalization capabilities by minimizing the model’s internal representations and salient tokens. |
Selecting Better Samples from Pre-trained LLMs: A Case Study on Question Generation (2023.findings-acl)
Copied to clipboard
Xingdi Yuan, Tong Wang, Yen-Hsiang Wang, Emery Fine, Rania Abdelghani, Hélène Sauzéon, Pierre-Yves Oudeyer
| Challenge: | Large Language Models (LLMs) have demonstrated impressive prowess in natural language generation. |
| Approach: | They propose a method to select high-quality questions from LLM-generated candidates using round-trip and prompt-based scoring. |
| Outcome: | The proposed approach can select high-quality questions from a set of LLM-generated candidates without modification of the underlying model nor rely on human annotations. |