Timo Schick (@timo_schick) / X

Timo Schick (@timo_schick) / X

🧩 How to Outperform GPT-3 by Combining Task Descriptions With Supervised Learning

Timo Schick on X: 🎥Missed the live stream? Make sure to check out this recording of @HinrichSchuetze's keynote Humans Learn From Task Descriptions and So Should Our Models at the first @ELLISforEurope

Michele Bevilacqua (@MicheleBevila20) / X

Timo Schick (@timo_schick) / X

Timo Schick on X: PET ( now runs with the latest version of @huggingface's transformers library. This means it is now possible to perform zero-shot and few-shot PET learning with multilingual models

Timo Schick (@timo_schick) / X

Diego Ardila (@Diego_Ardila1) / X

timo tons of chicks｜TikTok Search

Timo Schick (@timo_schick) / X

Timo Schick (@timo_schick) / X

AJAkil (@AJAkil2) / X

Ves Stoyanov (@vesko_st) / X

🧩 How to Outperform GPT-3 by Combining Task Descriptions With Supervised Learning

Exploiting Cloze-Questions for Few-Shot Text Classification and Natural Language Inference - ACL Anthology