My student Na Yeon Lee will be at #naacl2024 to present our paper on "Exploring Cross-Cultural Differences in English Hate Speech Annotations" arxiv.org/pdf/2308.16705 We start with the simple fact that English is widely spoken globally, but current NLP datasets are focused on how

Didn’t even cite the original work 🤦‍♂️

Another 'WOW' paper - Upto 20x improvement in inference throughput with Block Transformer compared to vanilla transformers with equivalent perplexity.🤯 How ❓ by MASSIVELY reducing KV cache IO overhead from quadratic to linear with respect to context length, solving a key

Has anyone tried few-shot prompting a base model with existing jailbreaking techniques? Could give us a few more papers.

EXAONE 3.0 7.8B beats Llama 3 8B and Gemma 2 9B on MT-Bench and more! Amazing work my previous collaborators and mentors at LG AI Research 🚀

Recent Apple Foundation models apply heavy quantization with LoRA to mitigate performance degradation, to fit 3B models on their devices. Check out the latest 🔬 details in QLoRA techniques from *the* efficiency expert 🧙‍♂️ from Korea.