WANDR
Benchmark for deep and wide research
Perplexity's benchmark and dataset for evaluating deep, wide research work, comprising 500 tasks represented as hierarchical, independently verifiable source-backed records.
Recent stories
0 linked stories
No linked stories yet.