Benchmark RadarAI BENCHMARK PROFILE
LigBench
LigBench is an automated evaluation benchmark for AI research idea generation, using fine-grained and reliable evaluation across generation distributions. It includes PAIR-IQ, a dataset for training pairwise idea judgment models.
- Released
- 2026-08-13
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Current evaluation of research idea generation is fragmented and lacks objective standards. LigBench aims to provide stable, interpretable, and expert-aligned assessments to support scalable and objective evaluation in this emerging area.
Motivation
With the rapid advancement of large language models (LLMs), research idea generation has attracted increasing attention.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.