Papers › Mini-Gemini: Mining the Potential of Multi-modality Vision Language Models

Mini-Gemini: Mining the Potential of Multi-modality Vision Language Models

27 Mar 2024arXiv:2403.18814archive 2025-07-28

Yanwei Li, Yuechen Zhang, Chengyao Wang, Zhisheng Zhong, Yixin Chen, Ruihang Chu, Shaoteng Liu, Jiaya Jia

In this work, we introduce Mini-Gemini, a simple and effective framework enhancing multi-modality Vision Language Models (VLMs). Despite the advancements in VLMs facilitating basic visual dialog and reasoning, a performance gap persists compared to advanced models like GPT-4 and Gemini. We try to narrow the gap by mining the potential of VLMs for better performance and any-to-any workflow from three aspects, i.e., high-resolution visual tokens, high-quality data, and VLM-guided generation. To enhance visual tokens, we propose to utilize an additional visual encoder for high-resolution refinement without increasing the visual token count. We further construct a high-quality dataset that promotes precise image comprehension and reasoning-based generation, expanding the operational scope of current VLMs. In general, Mini-Gemini further mines the potential of VLMs and empowers current frameworks with image understanding, reasoning, and generation simultaneously. Mini-Gemini supports a series of dense and MoE Large Language Models (LLMs) from 2B to 34B. It is demonstrated to achieve leading performance in several zero-shot benchmarks and even surpasses the developed private models. Code and models are available at https://github.com/dvlab-research/MiniGemini.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2403.18814")

Code

Syntology Ran 7 of 8 code samples harvested from 1 repository linked to this paper; 1 has no recorded run. Of those that ran: 1 ran · honoured contract; 1 ran · violated contract; 3 ran · our draft was wrong; 2 ran with no contract checked.

By repository: official repository: 8 samples from 1 repository, 7 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

dvlab-research/minigemini officialmentioned in papermentioned on GitHubpytorchApache-2.0 report
dvlab-research/MGM mentioned on GitHubpytorchApache-2.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

8 samples harvested; 7 ran; 1 honoured the contract we drafted; 1 has no recorded run. Read from Syntology's graph 2026-09-25; that is when this build read the record, not when the samples ran.

1ran · honoured contract
1ran · violated contract
3ran · our draft was wrong
2ran
1unverified

Licence: 0 of the 8 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from dvlab-research/minigemini. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

create_one_query dvlab-research/minigemini/mgm/eval/model_math_vista.py official repository ran Apache-2.0 (permissive) · 269bb26a4378dd51 · report
expand2square dvlab-research/minigemini/mgm/mm_utils.py official repository ran · our draft was wrong Apache-2.0 (permissive) · 592b3c1a88f93d7c · report
get_chunk dvlab-research/minigemini/mgm/eval/model_math_vista.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 42a46570620cd9fa · report
is_none dvlab-research/minigemini/mgm/eval/model_vqa_mmbench.py official repository ran · violated contract Apache-2.0 (permissive) · bae18947b56f2be1 · report
load_image dvlab-research/minigemini/predict.py official repository ran · honoured contract Apache-2.0 (permissive) · 9b3c1cb391672ccb · report
load_image_from_base64 dvlab-research/minigemini/mgm/mm_utils.py official repository ran Apache-2.0 (permissive) · c3ee9d07c900dd55 · report
split_list dvlab-research/minigemini/mgm/eval/model_math_vista.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 076c252c52cbb161 · report
process_images dvlab-research/minigemini/mgm/mm_utils.py official repository unverified Apache-2.0 (permissive) · d6153bc4456b4b0c · report

Tasks

Image ClassificationImage ComprehensionReferring Expression ComprehensionReferring expression generationVisual DialogVisual Question Answering

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ColonINST-v1 (Seen) MGM-2B (w/o LoRA, w/ extra data) Accuray 93.24 #7 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Seen) MGM-2B (w/o LoRA, w/o extra data) Accuray 92.97 #10 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Unseen) MGM-2B (w/o LoRA, w/o extra data) Accuray 78.99 #7 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Unseen) MGM-2B (w/o LoRA, w/ extra data) Accuray 78.69 #9 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Seen) MGM-2B (w/o LoRA, w/ extra data) Accuray 98.75 #4 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Seen) MGM-2B (w/o LoRA, w/o extra data) Accuray 98.17 #6 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Unseen) MGM-2B (w/o LoRA, w/ extra data) Accuray 74.30 #6 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Unseen) MGM-2B (w/o LoRA, w/o extra data) Accuray 69.81 #14 of 17 Archive leaderboard report
Visual Question Answering MM-Vet Mini-Gemini-HD-BS GPT-4 score 60.8 #35 of 231 Archive leaderboard report
Visual Question Answering MM-Vet Mini-Gemini-HD GPT-4 score 59.3 #40 of 231 Archive leaderboard report
Visual Question Answering MM-Vet Mini-Gemini GPT-4 score 53.0 #50 of 231 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutGPT-4Label SmoothingLayer NormalizationLinear LayerMoEMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections