Benchmark RadarIFCMemoryBench
IFCMemoryBench evaluates long-term memory in LLM-based agents for BIM information retrieval, with 143 multi-session tasks across 19 projects and 4,016 prior sessions, requiring integration of remembered context with live IFC queries.
- Released
- 2026-07-13
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Existing memory evaluations focus on conversational recall, not professional domains. IFCMemoryBench provides a test for whether agents can reuse information across sessions in a structured, domain-specific environment, revealing gaps in current memory systems.
Motivation
Long-term memory is becoming a core capability of LLM-based agents, but existing evaluations largely test conversational recall in open-domain or persona-grounded settings.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.