LLM Benchmarking Shows Capabilities Doubling Every 7 Months
Berkeley'deki METR araştırma enstitüsü, büyük dil modellerinin (LLM) yeteneklerinin her 7 ayda bir katlanarak arttığını ortaya koyan yeni bir değerlendirme yöntemi geliştirmiştir.
Researchers at the METR think tank have developed a new method for evaluating large language models (LLMs) by comparing their performance on tasks to human completion times. This approach revealed that LLM capabilities are improving exponentially, with their ability to reliably complete more complex tasks doubling approximately every seven months. Extrapolating this trend suggests that by 2030, advanced LLMs could potentially handle tasks equivalent to 167 human working hours.
This rapid, exponential improvement in LLM capabilities has significant implications for future technological development and the economy, potentially automating increasingly complex tasks.
📌 Kaynak
Bu özet spectrumieee kaynağından otomatik derlenmiştir. Tamamı için orijinal habere gidin.
Orijinal haberi oku →