Benchmark RadarProEvent
ProEvent is an event-centric benchmark for proactive agents, evaluating their ability to maintain a user's timetable from instant messaging chats. It assesses response timing, single-step correctness, and multi-step correctness using synthesized realistic chat scenarios.
- Released
- 2026-07-20
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
The benchmark fills a gap in evaluating proactive agents for event-centric assistance, which is crucial for autonomous support. It provides practical value in measuring agents' ability to detect implicit events and reason from the user's perspective, revealing significant limitations in current models.
Motivation
Proactive agents are expected to anticipate user needs and provide autonomous assistance by perceiving environmental context without explicit instructions.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.