LightOn Releases Multimodal Reranker
tomaarsen · x · 2026-07-17
LightOnIO has released LightOn-rerank: a set of rerankers designed to score text passages or document page images against a given query.
- A total of 6 models, available in 0.8B / 2B / 4B scales;
- Each scale offers both pointwise and generative listwise versions;
- Ideal for text ↔ image retrieval.
The author also notes that the 4B listwise model achieved 64.69 on ViDoRe V3 NDCG@10, slightly outperforming the standard 8B VL Reranker (64.23 under the same protocol) while using only half the parameters.
Related event: LightOn-rerank Targets Mixed-Corpus RAG With Multimodal Rerankers(10 posts)→
More from Research
- Programmable Cellular Automata: CA rules as readable code for explainability — Amidos2006 · 2026-09-11
- GEVIBench launches as a comprehensive benchmark for comparing voltage indicators — drmichaellevin · 2026-09-11
- Gaussian Light Transport: 13D Gaussian Mixtures Speed Up Global Illumination — ssh4net · 2026-09-11
- Llama Loves Pirates — Goodfire's Tom McGrath on teaching math without the pirate style — Machine Learning Street Talk · 2026-09-11
- Fortnow: P vs NP beyond AI's reach, but NP vs L separations could fall — fortnow · 2026-09-11
- MaP-WAM tackles non-Markovian robot manipulation with memory-grounded planning — Sizhe Zhao · 2026-09-11