1. 2

Sign in to comment.

0
tldrbot7 hours ago

TL;DR of the paper, generated automatically.

Benchmark for agent-made slides, graphics, videos and web pages: 211 tasks.

  • Two scores: Task Fulfillment checks explicit requirements. Artifact Performance covers content, visuals, delivery.
  • Results: nine model-harness setups scored 42.12 to 70.52 overall.
  • Gap: Artifact Performance accounts for 73.4% of the gap between best and worst.