Papers › Long Context Transfer from Language to Vision

Long Context Transfer from Language to Vision

24 Jun 2024arXiv:2406.16852archive 2025-07-28

Peiyuan Zhang, Kaichen Zhang, Bo Li, Guangtao Zeng, Jingkang Yang, Yuanhan Zhang, Ziyue Wang, Haoran Tan, Chunyuan Li, Ziwei Liu

Video sequences offer valuable temporal information, but existing large multimodal models (LMMs) fall short in understanding extremely long videos. Many works address this by reducing the number of visual tokens using visual resamplers. Alternatively, in this paper, we approach this problem from the perspective of the language model. By simply extrapolating the context length of the language backbone, we enable LMMs to comprehend orders of magnitude more visual tokens without any video training. We call this phenomenon long context transfer and carefully ablate its properties. To effectively measure LMMs' ability to generalize to long contexts in the vision modality, we develop V-NIAH (Visual Needle-In-A-Haystack), a purely synthetic long vision benchmark inspired by the language model's NIAH test. Our proposed Long Video Assistant (LongVA) can process 2000 frames or over 200K visual tokens without additional complexities. With its extended context length, LongVA achieves state-of-the-art performance on Video-MME among 7B-scale models by densely sampling more input frames. Our work is open-sourced at https://github.com/EvolvingLMMs-Lab/LongVA.

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="2406.16852")

Code

Syntology Ran 5 of 5 code samples harvested from 1 repository linked to this paper; 0 have no recorded run. Of those that ran: 2 ran · honoured contract; 3 ran · our draft was wrong.

By repository: official repository: 2 samples from 1 repository, 2 ran; 3 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

evolvinglmms-lab/longva officialmentioned in papermentioned on GitHubpytorchNOASSERTION report
jzhang38/EasyContext mentioned on GitHubpytorch 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

5 samples harvested; 5 ran; 2 honoured the contract we drafted; 0 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.

2ran · honoured contract
3ran · our draft was wrong

Licence: 5 of the 5 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 evolvinglmms-lab/longva. Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “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.

load_haystack evolvinglmms-lab/longva/vision_niah/eval_vision_niah.py official repository ran · our draft was wrong licence not identified · pointer only · 70759df34aae94c5 · report
safe_tokenize evolvinglmms-lab/longva/vision_niah/eval_vision_niah.py official repository ran · honoured contract licence not identified · pointer only · 300bb78b66d35115 · report
construct_prompt identical code first harvested elsewhere ran · our draft was wrong licence of this copy not recorded · 2e94fa2078ca5f7c · report
generate_file_hash identical code first harvested elsewhere ran · our draft was wrong licence of this copy not recorded · 08a5ceaa1b4682c8 · report
load_haystack identical code first harvested elsewhere ran · honoured contract licence of this copy not recorded · 83526663aea6f3a5 · report

Tasks

Language ModelingLanguage ModellingMMEVideo Question AnsweringVisual Question Answering (VQA)Zero-Shot Video Question Answer

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Video Question Answering OVBench LongVA (7B) AVG 43.6 #8 of 16 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B Average Score on VLM2-bench (9 subtasks) 22.59 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B GC-mat 14.29 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B GC-trk 19.18 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B OC-cnt 42.53 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B OC-cpr 26.67 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B OC-grp 18.50 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B PC-VID 3.75 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B PC-cnt 38.90 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B PC-cpr 21.50 #9 of 9 Archive leaderboard report
Visual Question Answering (VQA) VLM2-Bench LongVA-7B PC-grp 18.00 #9 of 9 Archive leaderboard report
Zero-Shot Video Question Answer NExT-QA LongVA(32 frames) Accuracy 67.1 #15 of 27 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.

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