Papers › Efficient Large Multi-modal Models via Visual Context Compression

Efficient Large Multi-modal Models via Visual Context Compression

28 Jun 2024arXiv:2406.20092archive 2025-07-28

Jieneng Chen, Luoxin Ye, Ju He, Zhao-Yang Wang, Daniel Khashabi, Alan Yuille

While significant advancements have been made in compressed representations for text embeddings in large language models (LLMs), the compression of visual tokens in multi-modal LLMs (MLLMs) has remained a largely overlooked area. In this work, we present the study on the analysis of redundancy concerning visual tokens and efficient training within these models. Our initial experiments show that eliminating up to 70% of visual tokens at the testing stage by simply average pooling only leads to a minimal 3% reduction in visual question answering accuracy on the GQA benchmark, indicating significant redundancy in visual context. Addressing this, we introduce Visual Context Compressor, which reduces the number of visual tokens to enhance training and inference efficiency without sacrificing performance. To minimize information loss caused by the compression on visual tokens while maintaining training efficiency, we develop LLaVolta as a light and staged training scheme that incorporates stage-wise visual context compression to progressively compress the visual tokens from heavily to lightly compression during training, yielding no loss of information when testing. Extensive experiments demonstrate that our approach enhances the performance of MLLMs in both image-language and video-language understanding, while also significantly cutting training costs and improving inference efficiency.

PaperPDFCodeCode Syntology ran

In Syntology 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="2406.20092")

Code

Syntology Ran 7 of 9 code samples harvested from 1 repository linked to this paper; 2 have no recorded run. Of those that ran: 1 ran · violated contract; 2 ran · our draft was wrong; 2 ran · fixture could not drive it; 2 ran with no contract checked.

By repository: official repository: 9 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.

beckschen/llavolta officialmentioned in paperpytorchApache-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

9 samples harvested; 7 ran; 0 honoured the contract we drafted; 2 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · violated contract
2ran · our draft was wrong
2ran · fixture could not drive it
2ran
2unverified

Licence: 0 of the 9 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 Beckschen/LLaVolta. “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.

divide_to_patches Beckschen/LLaVolta/llava/mm_utils.py official repository ran Apache-2.0 (permissive) · 7e03b180fa317c9a · report
get_chunk Beckschen/LLaVolta/llava/eval/model_vqa.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 42a46570620cd9fa · report
is_none Beckschen/LLaVolta/llava/eval/model_vqa_mmbench.py official repository ran · violated contract Apache-2.0 (permissive) · bae18947b56f2be1 · report
resize_and_pad_image Beckschen/LLaVolta/llava/mm_utils.py official repository ran Apache-2.0 (permissive) · 468eedeba67f1b00 · report
select_best_resolution Beckschen/LLaVolta/llava/mm_utils.py official repository ran · fixture could not drive it Apache-2.0 (permissive) · 3999ff487573f32c · report
split_list Beckschen/LLaVolta/llava/eval/model_vqa.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 076c252c52cbb161 · report
unpad_image Beckschen/LLaVolta/llava/model/llava_arch.py official repository ran · fixture could not drive it fingerprinted Apache-2.0 (permissive) · 7606525af238fb64 · report
DAbstractor beckschen/llavolta/llava/model/multimodal_projector/visual_plugin.py official repository unverified Apache-2.0 (permissive) · 03c055e43997ed75 · report
find_index_by_key Beckschen/LLaVolta/llava/eval/model_vqa_loader.py official repository unverified Apache-2.0 (permissive) · 8f1327d9a2b55348 · report

Tasks

Question AnsweringVisual Question Answering

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Visual Question Answering MM-Vet LLaVolta GPT-4 score 30.7 #203 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

Average Pooling

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