Papers › 3D-LLM: Injecting the 3D World into Large Language Models

3D-LLM: Injecting the 3D World into Large Language Models

24 Jul 2023NeurIPS 2023 11arXiv:2307.12981archive 2025-07-28

Yining Hong, Haoyu Zhen, Peihao Chen, Shuhong Zheng, Yilun Du, Zhenfang Chen, Chuang Gan

Large language models (LLMs) and Vision-Language Models (VLMs) have been proven to excel at multiple tasks, such as commonsense reasoning. Powerful as these models can be, they are not grounded in the 3D physical world, which involves richer concepts such as spatial relationships, affordances, physics, layout, and so on. In this work, we propose to inject the 3D world into large language models and introduce a whole new family of 3D-LLMs. Specifically, 3D-LLMs can take 3D point clouds and their features as input and perform a diverse set of 3D-related tasks, including captioning, dense captioning, 3D question answering, task decomposition, 3D grounding, 3D-assisted dialog, navigation, and so on. Using three types of prompting mechanisms that we design, we are able to collect over 300k 3D-language data covering these tasks. To efficiently train 3D-LLMs, we first utilize a 3D feature extractor that obtains 3D features from rendered multi- view images. Then, we use 2D VLMs as our backbones to train our 3D-LLMs. By introducing a 3D localization mechanism, 3D-LLMs can better capture 3D spatial information. Experiments on ScanQA show that our model outperforms state-of-the-art baselines by a large margin (e.g., the BLEU-1 score surpasses state-of-the-art score by 9%). Furthermore, experiments on our held-in datasets for 3D captioning, task composition, and 3D-assisted dialogue show that our model outperforms 2D VLMs. Qualitative examples also show that our model could perform more tasks beyond the scope of existing LLMs and VLMs. Project Page: : https://vis-www.cs.umass.edu/3dllm/.

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 10 code samples harvested from 2 repositories linked to this paper: 6 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 5 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 10 samples from 2 repositories, 6 ran (1 with no instrument failure; 5 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.

No sample is linked to this paper from 3 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-24 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 2 of the 6 samples that ran have no recorded date.

Pointcept/GPT4Point mentioned on GitHubpytorchMITSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
Yui010206/CREMA mentioned on GitHubpytorchBSD-3-ClauseSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to CREMA: Generalizable and Efficient Video-Language Reasoning via Multimodal Modular Fusion (arXiv:2402.05889; that paper's own run record: 8 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
openrobotlab/pointllm mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D Understanding,... (arXiv:2309.00615; that paper's own run record: 15 ran (of which 0 constructed an object rather than computing a result; 9 with no instrument failure: 3 honoured, 1 violated, 5 with no contract checked; 6 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. report
qizekun/ShapeLLM mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to ShapeLLM: Universal 3D Object Understanding for Embodied Interaction (arXiv:2402.17766; that paper's own run record: 14 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 5 honoured, 1 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 3 unverified); nothing checks that they implement this paper's method. report
umass-foundation-model/3d-llm mentioned on GitHubpytorchMITSyntology: 9 samples, 5 ran (5 instrument-failed) · 4 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.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

10 samples harvested: 6 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 5 where Syntology's instrument failed) · 4 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.

1ran · violated contract
5ran · our draft was wrong
4unverified

Licence: 10 of the 10 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 2 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. 1 sample whose link to this paper records no repository is listed under one of this paper's own repositories, because identical code was harvested from it for another paper. “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.

Pointcept/GPT4Point community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
disabled_train in 172 papers lavis/models/gpt4point_models/gpt4point.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran · violated contract licence of this copy not recorded · 4cb732f513d69dfd · report fixture: signature self:callable/0/?, mode:bool/0/bool · returned SimpleModel · level recorded, date not recorded
umass-foundation-model/3d-llm community (archive-listed) 9 samples, 5 ran (5 instrument-failed) · 4 unverified
compress in 35 papers
census: 1 implementation from 5 papers, 1 distinct output on one shared input
3DLanguage_data/ChatCaptioner_based/quetion_model.py ran · our draft was wrong MIT recorded; this copy not marked cleared · pointer only · 80dde39da1ba3708 · report fixture: signature tensor:Tensor/2/float, config:QuantizationConfig/0/? · returned tuple · level recorded 2026-09-25
interpolate_pos_embed in 59 papers 3DLLM_BLIP2-base/lavis/models/vit.py ran · our draft was wrong MIT recorded; this copy not marked cleared · pointer only · c6ec173f19f5c34d · report fixture: signature pos_embed_checkpoint:Tensor/3/float, visual_encoder:callable/0/? · returned Tensor · level recorded, date not recorded
processDigitArticle in 6 papers 3DLLM_BLIP2-base/calculate_scores/calculate_score_3dmvvqa.py ran · our draft was wrong fingerprinted MIT recorded; this copy not marked cleared · pointer only · 4e436e13094b20cb · report fixture: signature inText:str/0/str · returned str · level recorded 2026-09-24
processPunctuation in 6 papers 3DLLM_BLIP2-base/calculate_scores/calculate_score_3dmvvqa.py ran · our draft was wrong fingerprinted MIT recorded; this copy not marked cleared · pointer only · 35e84a8da1d4f427 · report fixture: signature inText:str/0/str · returned str · level recorded 2026-09-24
tile in 109 papers
census: 5 implementations from 7 papers, 4 distinct outputs across 3 buckets, one shared input per bucket
3DLLM_BLIP2-base/lavis/models/base_model.py ran · our draft was wrong fingerprinted MIT recorded; this copy not marked cleared · pointer only · cb33571427334815 · report fixture: signature x:Tensor/4/float, dim:int/0/int, n_tile:int/0/int · returned Tensor · level recorded 2026-09-24
all_gather_with_grad in 93 papers 3DLLM_BLIP2-base/lavis/models/base_model.py unverified MIT recorded; this copy not marked cleared · pointer only · 0ec9fc2025c16f65 · report
decompress in 6 papers 3DLanguage_data/ChatCaptioner_based/quetion_model.py unverified MIT recorded; this copy not marked cleared · pointer only · a2b79de775636d17 · report
get_gpu_memory in 16 papers 3DLanguage_data/ChatCaptioner_based/quetion_model.py unverified MIT recorded; this copy not marked cleared · pointer only · fe66d0280e436155 · report
get_lemma 3DLLM_BLIP2-base/calculate_scores/calculate_score_3dmvvqa.py unverified MIT recorded; this copy not marked cleared · pointer only · a1ff41c753bffe5d · report

Tasks

3D Object Captioning3D Question Answering (3D-QA)Dense CaptioningGenerative 3D Object ClassificationQuestion Answering

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Object Captioning Objaverse 3D-LLM Sentence-BERT 44.48 #6 of 6 Archive leaderboard report
3D Object Captioning Objaverse 3D-LLM Correctness 1.77 #6 of 6 Archive leaderboard report
3D Object Captioning Objaverse 3D-LLM GPT-4 33.42 #6 of 6 Archive leaderboard report
3D Object Captioning Objaverse 3D-LLM Hallucination 1.16 #6 of 6 Archive leaderboard report
3D Object Captioning Objaverse 3D-LLM Precision 60.39 #6 of 6 Archive leaderboard report
3D Object Captioning Objaverse 3D-LLM SimCSE 43.68 #6 of 6 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (flamingo) BLEU-1 32.6 #8 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (flamingo) BLEU-4 8.4 #8 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (flamingo) CIDEr 65.6 #8 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (flamingo) Exact Match 23.2 #8 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (flamingo) METEOR 13.5 #8 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (flamingo) ROUGE 34.8 #8 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-flant5) BLEU-1 38.3 #14 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-flant5) BLEU-4 11.6 #14 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-flant5) CIDEr 69.6 #14 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-flant5) Exact Match 19.1 #14 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-flant5) METEOR 14.9 #14 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-flant5) ROUGE 35.3 #14 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-opt) BLEU-1 37.3 #15 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-opt) BLEU-4 10.7 #15 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-opt) CIDEr 67.1 #15 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-opt) Exact Match 19.1 #15 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-opt) METEOR 14.3 #15 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects 3D-LLM (BLIP2-opt) ROUGE 34.5 #15 of 18 Archive leaderboard report
Generative 3D Object Classification Objaverse 3D-LLM Objaverse (Average) 45.25 #6 of 7 Archive leaderboard report
Generative 3D Object Classification Objaverse 3D-LLM Objaverse (C) 41.50 #6 of 7 Archive leaderboard report
Generative 3D Object Classification Objaverse 3D-LLM Objaverse (I) 49.00 #6 of 7 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