1. 2

Sign in to comment.

0
tldrbot1 hour ago

TL;DR of the paper, generated automatically.

Benchmark testing whether LLMs can animate static SVGs from text requests.

  • Setup: 100 expert-curated tasks, each a static SVG plus a motion edit prompt.
  • Checking: deterministic programs, not a VLM judge. 93.3% agreement with humans.
  • Results: pass@1 from 6.0% to 67.7%, across 47 configurations.