Papers › LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models

LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models

28 Nov 2023arXiv:2311.17043archive 2025-07-28

Yanwei Li, Chengyao Wang, Jiaya Jia

In this work, we present a novel method to tackle the token generation challenge in Vision Language Models (VLMs) for video and image understanding, called LLaMA-VID. Current VLMs, while proficient in tasks like image captioning and visual question answering, face computational burdens when processing long videos due to the excessive visual tokens. LLaMA-VID addresses this issue by representing each frame with two distinct tokens, namely context token and content token. The context token encodes the overall image context based on user input, whereas the content token encapsulates visual cues in each frame. This dual-token strategy significantly reduces the overload of long videos while preserving critical information. Generally, LLaMA-VID empowers existing frameworks to support hour-long videos and pushes their upper limit with an extra context token. It is proved to surpass previous methods on most of video- or image-based benchmarks. Code is available https://github.com/dvlab-research/LLaMA-VID}{https://github.com/dvlab-research/LLaMA-VID

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="2311.17043")

Code

Syntology Ran 3 of 4 code samples harvested from 1 repository linked to this paper; 1 has no recorded run. Of those that ran: 1 ran · violated contract; 2 ran · our draft was wrong.

By repository: official repository: 4 samples from 1 repository, 3 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/llama-vid officialmentioned in papermentioned 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

4 samples harvested; 3 ran; 0 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 · violated contract
2ran · our draft was wrong
1unverified

Licence: 0 of the 4 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/llama-vid. “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.

get_chunk dvlab-research/llama-vid/llamavid/eval/model_activitynet_qa.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 42a46570620cd9fa · report
is_none dvlab-research/llama-vid/llamavid/eval/model_vqa_mmbench.py official repository ran · violated contract Apache-2.0 (permissive) · bae18947b56f2be1 · report
split_list dvlab-research/llama-vid/llamavid/eval/model_activitynet_qa.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · 076c252c52cbb161 · report
apply_rotary_pos_emb_inference dvlab-research/llama-vid/llamavid/train/llama_flash_attn_monkey_patch.py official repository unverified Apache-2.0 (permissive) · d367874245639411 · report

Tasks

Image CaptioningQuestion AnsweringVideo Question AnsweringVideo-based Generative Performance BenchmarkingVisual Question AnsweringZero-Shot Video Question Answer

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Video Question Answering ActivityNet-QA LLaMA-VID-13B (2 Token) Accuracy 47.5 #11 of 36 Archive leaderboard report
Video Question Answering ActivityNet-QA LLaMA-VID-13B (2 Token) Confidence score 3.3 #11 of 36 Archive leaderboard report
Video Question Answering ActivityNet-QA LLaMA-VID-7B (2 Token) Accuracy 47.4 #12 of 36 Archive leaderboard report
Video Question Answering ActivityNet-QA LLaMA-VID-7B (2 Token) Confidence score 3.3 #12 of 36 Archive leaderboard report
Video Question Answering OVBench LLaMA-VID (7B) AVG 41.9 #9 of 16 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-13B (2 Token) Consistency 2.63 #13 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-13B (2 Token) Contextual Understanding 3.60 #13 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-13B (2 Token) Correctness of Information 3.07 #13 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-13B (2 Token) Detail Orientation 3.05 #13 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-13B (2 Token) Temporal Understanding 2.58 #13 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-13B (2 Token) mean 2.99 #13 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-7B (2 Token) Consistency 2.51 #16 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-7B (2 Token) Contextual Understanding 3.53 #16 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-7B (2 Token) Correctness of Information 2.96 #16 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-7B (2 Token) Detail Orientation 3.00 #16 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-7B (2 Token) Temporal Understanding 2.46 #16 of 23 Archive leaderboard report
Video-based Generative Performance Benchmarking VideoInstruct LLaMA-VID-7B (2 Token) mean 2.89 #16 of 23 Archive leaderboard report
Zero-Shot Video Question Answer ActivityNet-QA LLaMA-VID-13B (2 Token) Accuracy 47.5 #15 of 28 Archive leaderboard report
Zero-Shot Video Question Answer ActivityNet-QA LLaMA-VID-13B (2 Token) Confidence Score 3.3 #15 of 28 Archive leaderboard report
Zero-Shot Video Question Answer ActivityNet-QA LLaMA-VID-7B (2 Token) Accuracy 47.4 #16 of 28 Archive leaderboard report
Zero-Shot Video Question Answer ActivityNet-QA LLaMA-VID-7B (2 Token) Confidence Score 3.3 #16 of 28 Archive leaderboard report
Zero-Shot Video Question Answer MSRVTT-QA LLaMA-VID-13B (2 Token) Accuracy 58.9 #18 of 30 Archive leaderboard report
Zero-Shot Video Question Answer MSRVTT-QA LLaMA-VID-13B (2 Token) Confidence Score 3.3 #18 of 30 Archive leaderboard report
Zero-Shot Video Question Answer MSRVTT-QA LLaMA-VID-7B (2 Token) Accuracy 57.7 #19 of 30 Archive leaderboard report
Zero-Shot Video Question Answer MSRVTT-QA LLaMA-VID-7B (2 Token) Confidence Score 3.2 #19 of 30 Archive leaderboard report
Zero-Shot Video Question Answer MSVD-QA LLaMA-VID-13B (2 Token) Accuracy 70.0 #19 of 28 Archive leaderboard report
Zero-Shot Video Question Answer MSVD-QA LLaMA-VID-13B (2 Token) Confidence Score 3.7 #19 of 28 Archive leaderboard report
Zero-Shot Video Question Answer MSVD-QA LLaMA-VID-7B (2 Token) Accuracy 69.7 #20 of 28 Archive leaderboard report
Zero-Shot Video Question Answer MSVD-QA LLaMA-VID-7B (2 Token) Confidence Score 3.7 #20 of 28 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