Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models
<h2 class="text-2xl font-bold mb-4">Summary</h2>
This paper, from Google, introduces and analyzes the BIG-bench benchmark, a large-scale, diverse suite of tasks designed to evaluate the capabilities ...