LightOn Releases Strong Multilingual Retrieval Models with Zero-Shot Generalization
antoine_chaffin · x · 2026-07-31
LightOn has released new multilingual retrieval models that demonstrate very strong performance. The key highlight is that their multilingual training with late interaction successfully generalizes to languages unseen during contrastive training.
The models also extend their capabilities to long-context and code retrieval using a translate-train approach, pushing the frontier of open retrieval technologies.
Related event: LightOn Open-Sources Multilingual Retrieval Models mDenseOn and mLateOn(18 posts)→
More from Models
- Thrive Holdings benchmarks AI on real tax and IT work across 75+ firms — BorisMPower · 2026-10-16
- A Little Encouragement Makes a Small Local Model Draw a Stunning Pelican SVG — Fancy-Snow7 · 2026-10-16
- Rapid-MLX 0.16.2 brings NVIDIA Nemotron 3.5 up to 120B to Apple Silicon Macs — kuchaev · 2026-10-16
- "End of Chinese OSS" claim pushed back: open models get license almost automatically — eliebakouch · 2026-10-16
- Opus 5.5 still leads a month later, but open-weight decision models are closing in — airesearch12 · 2026-10-16
- Google restricts Gemini Pro access for Google Plus subscribers, matching free tier — EmergencyLess8861 · 2026-10-16