International Conference on Learning Representations (ICLR) 2021
Blog: Debiasing large pretrained language models using distributional control
Seminar: Perceptual robot learning
Continual adaptation of visual representations via data augmentation and meta-learning
Probabilistic embeddings for cross-modal retrieval
An empirical study of end-to-end simultaneous speech translation decoding strategies
Risk conditioned distributional soft actor-critic for risk-sensitive navigation
Debiasing large pretrained language models using distributional control
A novel framework for controlled NLG called 'Generation with Distributional Control', achieves great generality on the types of constraints that can be imposed and has a large potential to remedy the problem of bias in language models.
NAVER LABS Europe @ EACL2021
Papers and activities of NAVER LABS Europe at EACL 2021. NLG, scaling to larger context, cost of supervised data, controlling the output plus ASR at the AfricaNLP workshop.
A scalable Transformer architecture for summarizing long documents
Our Global BERT-based Transformer architecture fuses global and local information at every layer, resulting in a reading comprehension model that achieves a deeper understanding for long documents and enables flexibility for downstream tasks.
GLOBAL AI R&D BELT
ACADEMIA – EU/GOVT – ENTREPRENEURS
Our partnerships range from long-term fundamental research to investment in products and services.
