🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

My lab École normale supérieure | PSL is recruiting interns! In particular I'm looking for people to explore two topics: - Domain Discovery for Blind Distributional Robustness - Curriculum design for Emergent Communication More info (and other topics): cognitive-ml.fr/docs/callForIn…

GPT-4 has been out for 72 hours, and it could change the world! Here are some amazing and important things it *can't* do (yet) ⬇️

While ChatGPT and Bard are amazing NLP progress in English, under-represented languages often lack necessary data and evaluation to benefit from the technology advancement. Our benchmark focusing on under-represented languages is the first step towards closing the gap.

Despite the fantastic progress we've seen recently in cross-lingual modeling, the best systems still make a lot of factual errors. To address this, here is our work on 🚨 Evaluating and Modeling Attribution for Cross-Lingual Question Answering 🚨 #1 Attribution Evaluation: Our

We propose 𝗺𝗺𝗧𝟱 a modular multilingual seq2seq model. Our modular design and training regime solves source language hallucinations resulting in massive performance gains in cross-lingual transfer scenarios. 📄 arxiv.org/abs/2305.14224

FIAT: Fusing learning paradigms with Instruction-Accelerated Tuning paper page: huggingface.co/papers/2309.04… Learning paradigms for large language models (LLMs) currently tend to fall within either in-context learning (ICL) or full fine-tuning. Each of these comes with their own

Can we achieve effective zero-shot summarization by combining language & task information from existing PEFT adapters with weight arithmetic? Our paper, accepted at the MRL workshop #EMNLP2024, explores this question! A 🧵: