AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Magic-MM-Embedding: Towards Visual-Token-Efficient Universal Multimodal Embedding with MLLMs

Qi Li, Yanzhe Zhao, Yongxin Zhou, Yameng Wang, Yandong Yang, Yuanjia Zhou, Jinxiang Liu

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Training-free token reduction for MLLMs · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add visual token compression (64 vs 256 tokens) (ablation)InternVL3-2B embedder · MMEB+VisDoc average (warm-up)average score: 66.1 -> 65.4No clear effect
add stage 1 capability restoration (ablation)Magic-MM-Embedding 2B · MMEB (36 datasets) (1.9B)Precision@1: 64.7 -> 65.4Improved
add stage 2 contrastive pretraining (ablation)Magic-MM-Embedding 2B · MMEB (36 datasets) (1.9B)Precision@1: 67.1 -> 68.0Improved
add stage 3 task-aware finetuning (ablation)Magic-MM-Embedding 2B · MMEB (36 datasets) (1.9B)Precision@1: 65.4 -> 68.0Improved
add reranker at inference (ablation)Magic-MM-Embedding 2B · MMEBPrecision@1: 68.0 -> 70.2Improved
add reranker at inference (ablation)Magic-MM-Embedding 2B · VisDocNDCG@5: 72.1 -> 73.3Improved
use MLLM-judged hard negatives instead of rule-based (ablation)Magic-MM-Embedding 2B stage 3 · MMEB (36 datasets) (1.9B)Precision@1: 67.0 -> 68.0Improved
swap judge Qwen3-VL-8B for InternVL3-8B (ablation)Magic-MM-Embedding 2B stage 3 · MMEB (36 datasets) (1.9B)Precision@1: 68.0 -> 67.8No clear effect
use bilinear interpolation instead of adaptive average pooling (ablation)InternVL3-VTC 2B · VisDoc (short warm-up)NDCG@5: 63.8 -> 64.6Improved
use bilinear interpolation instead of learnable Conv2d (ablation)InternVL3-VTC 2B · MMEB (short warm-up)Precision@1: 35.1 -> 57.4Improved
add visual token compression to cut training time (ablation)InternVL3-2B · MMEBPrecision@1: 52h 43m 35s -> 24h 38m 28sImproved
use LoRA rank 16 instead of 32 (ablation)Magic-MM-Embedding 2B warm-up · MMEB+VisDoc averageaverage score: 65.1 -> 65.7Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy