Benchmark RadarAI BENCHMARK PROFILE
GateMem
GateMem evaluates memory governance in multi-principal shared-memory agents across medical, office, education, and household domains. It measures utility, access control, and active forgetting via checkpoints and a composite score.
- Released
- 2026-06-17
- Readiness
- Runnable
- Primary field
- Health & Life Sciences
Why it matters
Shared-memory agents are understudied, yet crucial for institutional deployments. GateMem addresses the need for evaluating governance capabilities beyond simple recall, informing development of reliable multi-user agents.
Motivation
Memory benchmarks for LLM agents largely assume single-user settings, leaving shared assistants for hospitals, workplaces, campuses, and households understudied.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.