Hcompany Releases NeoMME: Multimodal-Native Multilingual Encoder with Masked Discrete Diffusion
Hcompany · hf · 2026-09-03
Hcompany released NeoMME on Hugging Face, a single-tower multimodal-native multilingual foundation encoder. It uses small bidirectional encoders pretrained with masked discrete diffusion, achieving strong visual document retrieval and high compression of late-interaction embeddings for efficient fine-tuning and inference.
More from Research
- Researchers: AI labs have dramatically underinvested in action-only monitors as CoT oversight fades — maksym_andr · 2026-09-03
- Google Officially Details Planetary Prediction Engine for Autonomous Global Modeling — TheTuringPost · 2026-09-03
- Google's Planetary Prediction Engine Shrinks Weeks-Long Geospatial Modeling to Minutes — TheTuringPost · 2026-09-03
- AI rebuilds the Iliad's Catalogue of Ships as an interactive 3D map with 1,186 vessels — eldonredwards · 2026-09-03
- Jan Kulveit: predicting few-body interactions is hard, million-part systems get easier again — gleech · 2026-09-03
- Hacking llama.cpp to hot-swap knowledge into Qwen's Ngram PLE table — ortegaalfredo · 2026-09-03