Task Completion Horizon

concepts · 1 notes linked

Related: Metr · Large Language Models · AI Benchmarking

Notes

• Large Language Model Performance Doubles Every 7 Months — LLM task-completion capability doubles every seven months exponentially