From 028a3dd6dcc8b43bf147f934fbee89b7b606d7bd Mon Sep 17 00:00:00 2001 From: Xiang An Date: Mon, 2 Mar 2026 10:48:31 +0800 Subject: [PATCH] Update README.md --- profile/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/profile/README.md b/profile/README.md index 003cc59..86be7db 100644 --- a/profile/README.md +++ b/profile/README.md @@ -22,7 +22,7 @@ NEO Series: Native Vision-Language Models built from first principles. Rethinks 📄 [Paper](https://arxiv.org/abs/2510.14979) | 🤗 [Models](https://huggingface.co/collections/Paranioar/neo1-0-68f0db9cbac952be3eca7089) -### [OneVision-Encoder](https://github.com/EvolvingLMMs-Lab/OneVision-Encoder) ⭐ 269 `CVPR 2025` +### [OneVision-Encoder](https://github.com/EvolvingLMMs-Lab/OneVision-Encoder) ⭐ 269 A vision encoder designed around codec-aligned sparsity as a foundational principle for multimodal intelligence. Abandons uniform computation to selectively encode only 3.1%-25% of regions rich in signal entropy, consistently outperforming Qwen3-ViT and SigLIP2 across 16 image, video, and document understanding benchmarks despite using substantially fewer visual tokens.