Press Space to continue
Press Space to continue
Have the expensive model teach the cheap one. @vesko_st shows agents can build readable, editable classifiers that match or beat fine-tuned RoBERTa across six tasks.
vesko_st@vesko_st3h ago
Jev matches frontier models on classification at a fraction of the cost. The obvious follow-up: can an agent build and maintain classifiers? I tested on six tasks. Opus built classifiers matched or beat fine-tuned RoBERTa on all six. https://t.co/WcJhCIjzbF