AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Liquid: Language Models are Scalable and Unified Multi-Modal Generators

Junfeng Wu, Yi Jiang, Chuofan Ma, Yuliang Liu, Hengshuang Zhao, Zehuan Yuan, Song Bai, Xiang Bai

IJCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants · 93 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 10M more visual generation data (ablation)Liquid pretraining · VQAv2accuracy: 60.7 -> 64.5Improved
add 10M more visual generation data (ablation)Liquid pretraining · MMEscore: 909.6 -> 1066.8Improved
add 10M more visual understanding data (ablation)Liquid pretraining · gFIDFID: 19.9 -> 14.9Improved
add 10M more generation data (ablation)Liquid pretraining · gFIDFID: 19.9 -> 12.8Improved
add 10M more understanding data (ablation)Liquid pretraining · GenAI-Bench overallVQAScore: 0.60 -> 0.66Improved
train on mixed text and image data instead of text only (Gemma-7B) (ablation)Liquid · text benchmarks MMLUaccuracy: 56.0 vs 55.5; HellaSwag 76.1 vs 77.2Mixed
train on mixed multimodal data instead of text-only (Llama3-1B) (ablation)Liquid · HellaSwagaccuracy: 62.7 -> 55.0Hurt
train on mixed multimodal data instead of text-only (Qwen2.5-32B) (ablation)Liquid · HellaSwagaccuracy: 79.2 -> 79.7No clear effect
train on mixed multimodal data instead of text-only (Qwen2.5-7B) (ablation)Liquid · MMLUaccuracy: 64.9 -> 62.0Hurt
longer pretraining by one more epoch (ablation)Liquid Gemma-7B · VQAv2accuracy: 68.0 -> 71.3Improved
replace Chameleon VQVAE with semantically aligned UniTok (ablation)Liquid · MMEscore: 1119.3 -> 1448.0Improved
replace Chameleon VQVAE with semantically aligned UniTok (ablation)Liquid · TextVQAaccuracy: 42.4 -> 51.6Improved

Later papers that compared with this method

1 later accepted paper reports Liquid (7B) as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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