[KongchangAI]
Benchmark

context-bench

A proprietary contextual embedding evaluation benchmark released by vector database vendor turbopuffer. Its test data is not publicly disclosed, with the goal of avoiding training data contamination and providing a more accurate reflection of model generalization ability.

Timeline (last 90 days)

Oct 5

pplx-embed-v2-context-9b-preview has achieved state-of-the-art (SOTA) results on two benchmarks: ConTEB and context-bench.

Unverified50%
Oct 5

context-bench is a benchmark newly released and privately held by vector database vendor turbopuffer.

Unverified50%

All Facts (2)

Source Articles