3 entries with this tag
A deceptively simple insight: if you ask a model to 'think step by step,' it reasons better. Chain-of-Thought prompting showed that intermediate reasoning stepsānot just final answersāunlock a model's latent reasoning ability.
The paper that bridged pretraining and ChatGPT. Instruction tuning showed how a simple formatādescribing tasks as natural languageācould make models dramatically better at understanding and following what you ask them to do.
A beginner-friendly explanation of BERT (Bidirectional Encoder Representations from Transformers), the 2018 paper that taught AI to understand language by reading in both directions. Follow-up to our 'Attention Is All You Need' explainer.