Benchmark RadarAI BENCHMARK PROFILE
LIBERO-VIFO
LIBERO-VIFO is a benchmark for visual cue following in vision-language-action models, evaluating capability and safety across eight cue families. However, no public artifacts or official links are provided.
- Released
- 2026-08-18
- Readiness
- Paper only
- Primary field
- Robotics & Autonomous Systems
Why it matters
If released, it could inform the development of safer VLA systems, but current lack of access limits its practical value.
Motivation
Visual cues are increasingly adopted to guide robot learning, but whether Vision-Language-Action (VLA) models can reliably follow authorized cues while disregarding unauthorized ones remains unclear.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.