2h agoPerplexity's Bo Wang says PPLX 0.6B late-interaction embedding model achieves 53% accuracy on BrowseComp-Plus benchmark— A GPT-5 and pplx-late pipeline reached 85.78% accuracy.——2——Original postCS#793@CSHORTEN30OPBOBo|@BO_WANGBOfinally i managed to catch up the BrowseComp-Plus train with PPLX 0.6B (single/multi-vec) models:6:30 AM · May 27, 2026 View on XReposted byOK#160|@LATEINTERACTION
