Papers › VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos

VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos

29 May 2024CVPR 2025 1arXiv:2405.19209archive 2025-07-28

Ziyang Wang, Shoubin Yu, Elias Stengel-Eskin, Jaehong Yoon, Feng Cheng, Gedas Bertasius, Mohit Bansal

Long-form video understanding has been a challenging task due to the high redundancy in video data and the abundance of query-irrelevant information. To tackle this challenge, we propose VideoTree, a training-free framework which builds a query-adaptive and hierarchical video representation for LLM reasoning over long-form videos. First, VideoTree extracts query-relevant information from the input video through an iterative process, progressively refining the selection of keyframes based on their relevance to the query. Furthermore, VideoTree leverages the inherent hierarchical structure of long video data, which is often overlooked by existing LLM-based methods. Specifically, we incorporate multigranularity information into a tree-based representation, allowing VideoTree to extract query-relevant details from long videos in a coarse-to-fine manner. This enables the model to effectively handle a wide range of video queries with varying levels of detail. Finally, VideoTree aggregates the hierarchical query-relevant information within the tree structure and feeds it into an LLM reasoning model to answer the query. Our experiments show that our training-free method improves both reasoning accuracy and efficiency compared to existing methods. Specifically, VideoTree outperforms the existing training-free approaches on the popular EgoSchema and NExT-QA benchmarks with less inference time, achieving 61.1% and 75.6% accuracy on the test set without additional video-specific training. Moreover, on the long split of Video-MME benchmark (average 44 minutes), the training-free VideoTree framework achieves better performance than the strong proprietary GPT-4V model and other MLLMs that were extensively trained on video data.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 11 code samples harvested from 1 repository linked to this paper: 11 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 2 honoured, 0 violated, 5 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified. Where Syntology's instrument failed: 4 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 11 samples from 1 repository, 11 ran (7 with no instrument failure; 4 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

Ziyang412/VideoTree officialmentioned on GitHubpytorchMITSyntology: 11 samples, 11 ran (4 instrument-failed) · 0 unverified 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

11 samples harvested: 11 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 2 honoured, 0 violated, 5 with no contract checked; 4 where Syntology's instrument failed) · 0 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · honoured contract
4ran · our draft was wrong
5ran

Licence: 0 of the 11 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 Ziyang412/VideoTree. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

Ziyang412/VideoTree official repository 11 samples, 11 ran (4 instrument-failed) · 0 unverified
build_fewshot_examples util.py ran fingerprinted MIT (permissive) · a16cc02636239688 · report fixture: signature qa_path:str/0/str, data_path:str/0/str · returned str · level recorded 2026-09-24
cosine_similarity in 120 papers
census: 26 implementations from 30 papers, 11 distinct outputs across 5 buckets, one shared input per bucket
depth_expansion.py ran · our draft was wrong fingerprinted MIT (permissive) · b92a31685a22299f · report fixture: signature points:Tensor/2/float, centroid:Tensor/1/float · returned Tensor · level recorded 2026-09-24
find_closest_points_in_temporal_order_subsub depth_expansion.py ran MIT (permissive) · 863db1147f3fdd20 · report fixture: signature x:Tensor/2/float, clusters:dict/0/?, relevance_scores:list/1/int · returned list · level recorded 2026-09-24
find_closest_points_per_cluster adaptive_breath_expansion.py ran MIT (permissive) · 4cd91b167b647601 · report fixture: signature x:Tensor/2/float, cluster_ids:Tensor/1/int, cluster_centers:Tensor/2/float · returned dict · level recorded 2026-09-24
first_char_after_anchor prompts.py ran · our draft was wrong MIT (permissive) · 3c2914cfae3c5c3a · report fixture: signature anchor:str/0/str · returned function · level recorded 2026-09-24
first_char_as_answer in 5 papers prompts.py ran · honoured contract fingerprinted MIT (permissive) · dd8763aca4a7d6b7 · report fixture: signature res:str/0/str · returned int · level recorded 2026-09-24
hierarchical_clustering_with_external_primary depth_expansion.py ran MIT (permissive) · e561f95df17f4c8b · report fixture: signature video_features:ndarray/2/float, cluster_ids:ndarray/1/int, relevance_scores:ndarray/1/int, num_subclusters:int/0/int, num_subsubclusters:int/0/int · returned dict · level recorded 2026-09-24
identity in 174 papers
census: 11 implementations from 39 papers, 3 distinct outputs across 2 buckets, one shared input per bucket
prompts.py ran · honoured contract fingerprinted MIT (permissive) · 036725c05a3d96f8 · report fixture: signature res:Tensor/2/float · returned Tensor · level recorded 2026-09-24
load_frame_features adaptive_breath_expansion.py ran MIT (permissive) · ec6d415be2b2d979 · report fixture: signature name_ids:str/0/str, save_folder:str/0/str · returned Tensor · level recorded 2026-09-24
load_json in 344 papers util.py ran · our draft was wrong MIT (permissive) · 658682feda4928a2 · report fixture: signature fn:str/0/str · returned dict · level recorded 2026-09-24
load_pkl in 46 papers util.py ran · our draft was wrong MIT (permissive) · 8080055de6293a6b · report fixture: signature fn:str/0/str · returned dict · level recorded 2026-09-24

Tasks

MMEVideo UnderstandingZero-Shot Video Question Answer

2 archive task tags without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Zero-Shot Video Question Answer EgoSchema (fullset) VideoTree (GPT4) Accuracy 61.1 #9 of 29 Archive leaderboard report
Zero-Shot Video Question Answer EgoSchema (subset) VideoTree (GPT4) Accuracy 66.2 #4 of 14 Archive leaderboard report
Zero-Shot Video Question Answer IntentQA VideoTree (GPT4) Accuracy 66.9 #5 of 13 Archive leaderboard report
Zero-Shot Video Question Answer NExT-QA VideoTree (GPT4) Accuracy 73.5 #6 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.

Methods

SET

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