Papers › Visual Instruction Tuning

Visual Instruction Tuning

17 Apr 2023NeurIPS 2023 11arXiv:2304.08485archive 2025-07-28

Haotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae Lee

Instruction tuning large language models (LLMs) using machine-generated instruction-following data has improved zero-shot capabilities on new tasks, but the idea is less explored in the multimodal field. In this paper, we present the first attempt to use language-only GPT-4 to generate multimodal language-image instruction-following data. By instruction tuning on such generated data, we introduce LLaVA: Large Language and Vision Assistant, an end-to-end trained large multimodal model that connects a vision encoder and LLM for general-purpose visual and language understanding.Our early experiments show that LLaVA demonstrates impressive multimodel chat abilities, sometimes exhibiting the behaviors of multimodal GPT-4 on unseen images/instructions, and yields a 85.1% relative score compared with GPT-4 on a synthetic multimodal instruction-following dataset. When fine-tuned on Science QA, the synergy of LLaVA and GPT-4 achieves a new state-of-the-art accuracy of 92.53%. We make GPT-4 generated visual instruction tuning data, our model and code base publicly available.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

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 51 code samples harvested from 12 repositories linked to this paper: 16 ran (of which 6 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 8 where Syntology's instrument failed) · 35 unverified. Where Syntology's instrument failed: 4 where our contract draft was wrong; 4 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 12 samples from 1 repository, 4 ran (1 with no instrument failure; 3 where Syntology's instrument failed); named in the paper: 5 samples from 1 repository, 2 ran (1 with no instrument failure; 1 where Syntology's instrument failed); community (archive-listed): 29 samples from 7 repositories, 9 ran (6 with no instrument failure; 3 where Syntology's instrument failed); community: 5 samples from 3 repositories, 1 ran (0 with no instrument failure; 1 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 2 of its repositories that Syntology harvested for other papers:

Levels written on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 6 of the 16 samples that ran have no recorded date.

13 repositories listed; official and paper-mentioned ones first.

haotian-liu/LLaVA officialmentioned on GitHubpytorchSyntology: 12 samples, 4 ran (3 instrument-failed) · 8 unverified report
llava-annonymous/llava mentioned in papermentioned on GitHubpytorchSyntology: 5 samples, 2 ran (1 instrument-failed) · 3 unverified report
computer-vision-in-the-wild/cvinw_readings mentioned in paperSyntology: not harvested report
LLaVA-VL/LLaVA-NeXT mentioned on GitHubpytorchSyntology: 3 samples, 2 ran (1 instrument-failed) · 1 unverified report
ZhangYiqun018/StickerConv mentioned on GitHubpytorchSyntology: 5 samples, 0 ran · 5 unverified report
camenduru/llava-colab mentioned on GitHubSyntology: not harvested report
dinhvietcuong1996/icme25-inova mentioned on GitHubpytorchSyntology: 12 samples, 5 ran (2 instrument-failed) · 7 unverified report
huggingface/transformers mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Attention Is All You Need (arXiv:1706.03762; that paper's own run record: 610 ran (of which 293 constructed an object rather than computing a result; 529 with no instrument failure: 45 honoured, 15 violated, 469 with no contract checked; 81 where Syntology's instrument failed) · 336 unverified), HuggingFace's Transformers: State-of-the-art Natural Language Processing (arXiv:1910.03771; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified), DeBERTa: Decoding-enhanced BERT with Disentangled Attention (arXiv:2006.03654; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 1 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 9 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
qiujihao19/artemis mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to GPT4RoI: Instruction Tuning Large Language Model on Region-of-Interest (arXiv:2307.03601; that paper's own run record: 3 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
skunkworksai/bakllava mentioned on GitHubpytorchSyntology: 3 samples, 0 ran · 3 unverified report
sshh12/multi_token mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
sunsmarterjie/chatterbox mentioned on GitHubpytorchSyntology: 3 samples, 1 ran (0 instrument-failed) · 2 unverified report
tabtoyou/kollava mentioned on GitHubpytorchSyntology: 2 samples, 0 ran · 2 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

51 samples harvested: 16 ran (of which 6 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 8 where Syntology's instrument failed) · 35 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · violated contract
4ran · our draft was wrong
4ran; the contract check could not run
7ran
35unverified

Licence: 0 of the 51 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 12 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. “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”. For 2 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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.

haotian-liu/LLaVA official repository 12 samples, 4 ran (3 instrument-failed) · 8 unverified
LlavaConfig in 6 papers llava/model/language_model/llava_llama.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · f01242335a1498b3 · report fixture: returned LlavaConfig · level recorded 2026-09-22
get_anyres_image_grid_shape in 6 papers llava/model/language_model/llava_llama.py ran; the contract check could not run (RAISES) · metamorphic tier: well formed Apache-2.0 (permissive) · 2ff0d742a84cb8e2 · report fixture: signature image_size:tuple/1/int, grid_pinpoints:list/1/tuple, patch_size:int/0/int · returned tuple · level recorded 2026-09-22
split_to_even_chunks in 41 papers llava/train/llava_trainer.py ran; the contract check could not run (RAISES) · metamorphic tier: well formed Apache-2.0 (permissive) · 10893c4608c08075 · report fixture: signature indices:list/1/int, lengths:dict/0/?, num_chunks:int/0/int · returned list · level recorded, date not recorded
unpad_image in 72 papers
census: 4 implementations from 34 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 3, kind float, dtype float32
llava/model/llava_arch.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · 55c32993da87759b · report fixture: signature tensor:Tensor/3/float, original_size:tuple/1/int · returned Tensor · level recorded, date not recorded
CLIPVisionTowerS2 llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 5d1c9ab7042cee8d · report
LlavaLlamaForCausalLM in 15 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 4363aef6226bb213 · report
LlavaLlamaModel in 13 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 9707b7b56a48db85 · report
LlavaMetaForCausalLM in 16 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 8e3647258cae7507 · report
LlavaMetaModel in 16 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 0bfdaed6a43a5263 · report
build_vision_tower in 35 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 3ee345b55cc88a11 · report
get_mm_adapter_state_maybe_zero_3 in 44 papers llava/train/llava_trainer.py unverified Apache-2.0 (permissive) · bb35e3ac741bb2c9 · report
maybe_zero_3 in 74 papers llava/train/llava_trainer.py unverified Apache-2.0 (permissive) · 735025744c1ab0cf · report
llava-annonymous/llava named in the paper 5 samples, 2 ran (1 instrument-failed) · 3 unverified
LlavaConfig in 6 papers llava/model/llava.py ran Apache-2.0 (permissive) · 92cb18f9e27a4e39 · report fixture: returned LlavaConfig · level recorded 2026-09-22
convert_caps llava/eval/eval_science_qa.py ran · our draft was wrong Apache-2.0 (permissive) · 94fda0761ce0d5b9 · report fixture: signature results:list/1/dict · returned list · level recorded, date not recorded
LlavaLlamaForCausalLM in 15 papers llava/model/llava.py unverified Apache-2.0 (permissive) · 7adeedb5a7dcbd7a · report
LlavaLlamaModel in 13 papers llava/model/llava.py unverified Apache-2.0 (permissive) · e992c05c75e7c542 · report
get_pred_idx llava/eval/eval_science_qa.py unverified Apache-2.0 (permissive) · 268b6194c5c23103 · report
LLaVA-VL/LLaVA-NeXT community (archive-listed) 3 samples, 2 ran (1 instrument-failed) · 1 unverified
should_only_save_mm_adapter llava/train/llava_trainer.py ran · violated contract Apache-2.0 (permissive) · 08787a49ef19e0aa · report fixture: signature args:Args/0/? · returned bool · level recorded, date not recorded
unpad_image in 72 papers
census: 4 implementations from 34 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 3, kind float, dtype float32
llava/model/llava_arch.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · 7606525af238fb64 · report fixture: signature tensor:Tensor/3/float, original_size:tuple/1/int · returned Tensor · level recorded, date not recorded
LlavaMetaForCausalLM in 16 papers llava/model/llava_arch.py unverified Apache-2.0 (permissive) · 39343c26374462af · report
ZhangYiqun018/StickerConv community (archive-listed) 5 samples, 0 ran · 5 unverified
LlavaLlamaForCausalLM in 15 papers sticker_process/llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 44969b8364ff17f1 · report
LlavaLlamaModel in 13 papers sticker_process/llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 7fc60199325c39f5 · report
LlavaMetaForCausalLM in 16 papers sticker_process/llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · abf7b5851a13bc76 · report
LlavaMetaModel in 16 papers sticker_process/llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 38d280bea74c1e01 · report
build_vision_tower in 35 papers sticker_process/llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 014050859277825c · report
dinhvietcuong1996/icme25-inova community (archive-listed) 12 samples, 5 ran (2 instrument-failed) · 7 unverified
IdentityMap llava/model/language_model/llava_llama.py ran · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · 1707c6987676247e · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
LlavaConfig in 6 papers llava/model/language_model/llava_llama.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · dbe52eb96f05b6db · report fixture: returned LlavaConfig · level recorded 2026-09-22
SimpleResBlock llava/model/language_model/llava_llama.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · 4a7e5ec29fb79acb · report fixture: returned SimpleResBlock · level recorded 2026-09-22
build_vision_projector in 23 papers llava/model/language_model/llava_llama.py ran · our draft was wrong Apache-2.0 (permissive) · ea7ba84c548c5a33 · report fixture: signature config:MockConfig/0/?, delay_load:bool/0/bool, vision_cfg:MockVisionConfig/0/? · returned Linear · level recorded 2026-09-22
build_vision_resampler llava/model/language_model/llava_llama.py ran · our draft was wrong Apache-2.0 (permissive) · 10f2142755f4bbb5 · report fixture: signature model_args:MockModelArgs/0/?, delay_load:bool/0/bool · returned IdentityMap · level recorded 2026-09-22
CLIPVisionTower in 11 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 98e6fcd805f498d1 · report
CLIPVisionTowerS2 llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · d7be74cc1bd52de7 · report
LlavaLlamaForCausalLM in 15 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 1717d5d0a430f17f · report
LlavaLlamaModel in 13 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · a39a4ca102cca6cc · report
LlavaMetaForCausalLM in 16 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 31bedbeb8856618e · report
LlavaMetaModel in 16 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 7732a5ae006ababd · report
build_vision_tower in 35 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 875ebbe8f3a0c711 · report
skunkworksai/bakllava community (archive-listed) 3 samples, 0 ran · 3 unverified
LlavaLlamaForCausalLM in 15 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · e589073e81d58266 · report
LlavaLlamaModel in 13 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · d9cce1fb106ab082 · report
LlavaMetaModel in 16 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 5990c8af59a1551f · report
sshh12/multi_token community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
_MLPVectorProjector multi_token/modalities/projectors.py ran fingerprinted Apache-2.0 (permissive) · 078a686c14a5d016 · report fixture: returned _MLPVectorProjector · level recorded 2026-09-22
sunsmarterjie/chatterbox community (archive-listed) 3 samples, 1 ran (0 instrument-failed) · 2 unverified
LlavaLlamaModel in 13 papers model/llava/model/llava.py ran Apache-2.0 (permissive) · e6460881f1cf311c · report fixture: returned LlavaLlamaModel · level recorded 2026-09-22
LlavaConfig in 6 papers model/llava/model/llava.py unverified Apache-2.0 (permissive) · d2904a880e39bfc2 · report
LlavaLlamaForCausalLM in 15 papers model/llava/model/llava.py unverified Apache-2.0 (permissive) · d138a906d3799efb · report
tabtoyou/kollava community (archive-listed) 2 samples, 0 ran · 2 unverified
LlavaLlamaForCausalLM in 15 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · d70c330590b6be51 · report
LlavaMetaForCausalLM in 16 papers llava/model/language_model/llava_llama.py unverified Apache-2.0 (permissive) · 2efb2d959762f4ed · report
ictnlp/LLaVA-Mini community 1 sample, 1 ran (1 instrument-failed) · 0 unverified
image_parser in 19 papers llavamini/eval/run_llava_mini.py ran · our draft was wrong Apache-2.0 (permissive) · a7bee88c1c7fd6a3 · report fixture: signature args:Args/0/? · returned list · level recorded, date not recorded
LLaVA-VL/LLaVA-Plus-Codebase community 2 samples, 0 ran · 2 unverified
convert_caps llava/eval/eval_science_qa.py unverified Apache-2.0 (permissive) · 150a20daa039f495 · report
get_pred_idx llava/eval/eval_science_qa.py unverified Apache-2.0 (permissive) · 63959934f6d1fa06 · report
PKU-YuanGroup/MoE-LLaVA community 2 samples, 0 ran · 2 unverified
save_image_to_local in 8 papers moellava/serve/gradio_web_server.py unverified Apache-2.0 (permissive) · e2e5c3a95a6aebb4 · report
save_video_to_local in 8 papers moellava/serve/gradio_web_server.py unverified Apache-2.0 (permissive) · 22930d58d32e2d08 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another.

9 assemblies on this paper: 9 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).

9 sibling assemblies for LLaVA (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for LLaVA (graph method: LLaVA). 5 bricks; individually on a fixture: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaMetaForCausalLM entry · level 0 · unverified · 39343c26374462af
  2. get_anyres_image_grid_shape helper · level 1 · ran; the contract check could not run · dd90994dc116b84b
  3. rank0_print helper · level 0 · unverified · 16c9ecf6f7fea719
  4. unpad_image helper · level 1 · ran; the contract check could not run (RAISES) · 7606525af238fb64
  5. select_best_resolution helper · level 1 · ran; the contract check could not run · 3999ff487573f32c

harvested Assembly for LLaVA (graph method: LLaVA). 16 bricks; individually on a fixture: 8 ran (of which 2 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 5 where Syntology's instrument failed) · 8 unverified. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · 1717d5d0a430f17f
  2. LlavaConfig helper · level 1 · ran · dbe52eb96f05b6db
  3. LlavaLlamaModel helper · level 0 · unverified · a39a4ca102cca6cc
  4. LlavaMetaForCausalLM helper · level 0 · unverified · 31bedbeb8856618e
  5. LlavaMetaModel model · level 0 · unverified · 7732a5ae006ababd
  6. get_anyres_image_grid_shape helper · level 1 · ran; the contract check could not run · dd90994dc116b84b
  7. rank0_print helper · level 0 · unverified · 16c9ecf6f7fea719
  8. unpad_image helper · level 1 · ran; the contract check could not run (RAISES) · 7606525af238fb64
  9. build_vision_projector helper · level 1 · ran · our draft was wrong · ea7ba84c548c5a33
  10. build_vision_resampler helper · level 1 · ran · our draft was wrong · 10f2142755f4bbb5
  11. build_vision_tower model · level 0 · unverified · 875ebbe8f3a0c711
  12. select_best_resolution helper · level 1 · ran; the contract check could not run · 3999ff487573f32c
  13. IdentityMap helper · level 1 · ran · 1707c6987676247e
  14. SimpleResBlock helper · level 1 · ran · 4a7e5ec29fb79acb
  15. CLIPVisionTower helper · level 0 · unverified · 98e6fcd805f498d1
  16. CLIPVisionTowerS2 helper · level 0 · unverified · d7be74cc1bd52de7

harvested Assembly for LLaVA (graph method: LLaVA). 9 bricks; individually on a fixture: 3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 6 unverified. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · 44969b8364ff17f1
  2. LlavaConfig helper · level 1 · ran · e9cfb7ae423d3e3a
  3. LlavaLlamaModel model · level 0 · unverified · 7fc60199325c39f5
  4. LlavaMetaForCausalLM model · level 0 · unverified · abf7b5851a13bc76
  5. LlavaMetaModel model · level 0 · unverified · 38d280bea74c1e01
  6. build_vision_projector helper · level 1 · ran · our draft was wrong · ef4db0da176c674c
  7. build_vision_tower helper · level 0 · unverified · 014050859277825c
  8. IdentityMap helper · level 1 · ran · 9f0f5d62445adabe
  9. CLIPVisionTower model · level 0 · unverified · 6422a6162a5e6169

harvested Assembly for LLaVA (graph method: LLaVA). 13 bricks; individually on a fixture: 7 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 4 where Syntology's instrument failed) · 6 unverified. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · 4363aef6226bb213
  2. LlavaConfig helper · level 1 · ran · f01242335a1498b3
  3. LlavaLlamaModel model · level 0 · unverified · 9707b7b56a48db85
  4. LlavaMetaForCausalLM helper · level 0 · unverified · 8e3647258cae7507
  5. LlavaMetaModel helper · level 0 · unverified · 0bfdaed6a43a5263
  6. get_anyres_image_grid_shape helper · level 1 · ran; the contract check could not run (RAISES) · 2ff0d742a84cb8e2
  7. unpad_image helper · level 1 · ran; the contract check could not run (RAISES) · 55c32993da87759b
  8. build_vision_projector model · level 1 · ran · our draft was wrong · ef4db0da176c674c
  9. build_vision_tower helper · level 0 · unverified · 3ee345b55cc88a11
  10. select_best_resolution helper · level 1 · ran; the contract check could not run · 3999ff487573f32c
  11. IdentityMap helper · level 1 · ran · 9f0f5d62445adabe
  12. CLIPVisionTower model · level 1 · ran · ee3792b1ae45cbb6
  13. CLIPVisionTowerS2 helper · level 0 · unverified · 5d1c9ab7042cee8d

harvested Assembly for LLaVA (graph method: LLaVA). 3 bricks; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · 7adeedb5a7dcbd7a
  2. LlavaConfig helper · level 1 · ran · 92cb18f9e27a4e39
  3. LlavaLlamaModel model · level 0 · unverified · e992c05c75e7c542

harvested Assembly for LLaVA (graph method: LLaVA). 9 bricks; individually on a fixture: 3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 6 unverified. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · e589073e81d58266
  2. LlavaConfig helper · level 1 · ran · e9cfb7ae423d3e3a
  3. LlavaLlamaModel model · level 0 · unverified · d9cce1fb106ab082
  4. LlavaMetaForCausalLM model · level 0 · unverified · abf7b5851a13bc76
  5. LlavaMetaModel helper · level 0 · unverified · 5990c8af59a1551f
  6. build_vision_projector helper · level 1 · ran · our draft was wrong · ef4db0da176c674c
  7. build_vision_tower helper · level 0 · unverified · 014050859277825c
  8. IdentityMap helper · level 1 · ran · 9f0f5d62445adabe
  9. CLIPVisionTower model · level 0 · unverified · 6422a6162a5e6169

harvested Assembly for LLaVA (graph method: LLaVA). 1 brick; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 9 siblings

  1. _MLPVectorProjector entry · level 1 · ran · 078a686c14a5d016

harvested Assembly for LLaVA (graph method: LLaVA). 3 bricks; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · d138a906d3799efb
  2. LlavaConfig helper · level 0 · unverified · d2904a880e39bfc2
  3. LlavaLlamaModel model · level 1 · ran · e6460881f1cf311c

harvested Assembly for LLaVA (graph method: LLaVA). 9 bricks; individually on a fixture: 3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 6 unverified. The assembly has not been run end-to-end.

one of 9 siblings

  1. LlavaLlamaForCausalLM entry · level 0 · unverified · d70c330590b6be51
  2. LlavaConfig helper · level 1 · ran · e9cfb7ae423d3e3a
  3. LlavaLlamaModel model · level 0 · unverified · 8fe347da378cfb91
  4. LlavaMetaForCausalLM helper · level 0 · unverified · 2efb2d959762f4ed
  5. LlavaMetaModel helper · level 0 · unverified · 2f7959e2c2e658be
  6. build_vision_projector helper · level 1 · ran · our draft was wrong · ef4db0da176c674c
  7. build_vision_tower helper · level 0 · unverified · 014050859277825c
  8. IdentityMap helper · level 1 · ran · 9f0f5d62445adabe
  9. CLIPVisionTower helper · level 0 · unverified · 6422a6162a5e6169

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2304.08485") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

1 Image, 2*2 Stitching3D Question Answering (3D-QA)Image ClassificationImage RetrievalInstruction FollowingMMR totalReferring Expression ComprehensionReferring expression generationSpatial ReasoningVideo Question AnsweringVisual Question AnsweringVisual Reasoningvisual instruction following

Datasets

Introduced by this paper, per the archive.

LLaVA-Bench

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Question Answering (3D-QA) ScanQA Test w/ objects LL3DA BLEU-1 - #18 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects LL3DA BLEU-4 13.5 #18 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects LL3DA CIDEr 76.8 #18 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects LL3DA Exact Match - #18 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects LL3DA METEOR 15.9 #18 of 18 Archive leaderboard report
3D Question Answering (3D-QA) ScanQA Test w/ objects LL3DA ROUGE 37.3 #18 of 18 Archive leaderboard report
Image Classification ColonINST-v1 (Seen) LLaVA-v1 (w/ LoRA, w/ extra data) Accuray 89.61 #15 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Seen) LLaVA-v1 (w/ LoRA, w/o extra data) Accuray 87.86 #16 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Unseen) LLaVA-v1 (w/ LoRA, w/o extra data) Accuray 72.08 #15 of 17 Archive leaderboard report
Image Classification ColonINST-v1 (Unseen) LLaVA-v1 (w/ LoRA, w/ extra data) Accuray 42.17 #17 of 17 Archive leaderboard report
MMR total MRR-Benchmark LLaVA-NEXT-34B Total Column Score 412 #4 of 14 Archive leaderboard report
MMR total MRR-Benchmark LLaVA-NEXT-13B Total Column Score 335 #8 of 14 Archive leaderboard report
MMR total MRR-Benchmark LLaVA-1.5-13B Total Column Score 243 #11 of 14 Archive leaderboard report
Referring expression generation ColonINST-v1 (Seen) LLaVA-v1 (w/ LoRA, w/ extra data) Accuray 86.87 #16 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Seen) LLaVA-v1 (w/ LoRA, w/o extra data) Accuray 84.55 #17 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Unseen) LLaVA-v1 (w/ LoRA, w/o extra data) Accuray 68.11 #16 of 17 Archive leaderboard report
Referring expression generation ColonINST-v1 (Unseen) LLaVA-v1 (w/ LoRA, w/ extra data) Accuray 46.85 #17 of 17 Archive leaderboard report
Spatial Reasoning EmbSpatial-Bench LLaVA-1.6 Generation 35.19 #4 of 5 Archive leaderboard report
Video Question Answering MVBench LLaVa Avg. 36.0 #17 of 22 Archive leaderboard report
Visual Question Answering BenchLMM LLaVA-1.5-7B GPT-3.5 score 46.83 #4 of 10 Archive leaderboard report
Visual Question Answering BenchLMM LLaVA-1-13B GPT-3.5 score 43.50 #7 of 10 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

Absolute Position EncodingsAdamAttentionBASEBPEDense ConnectionsDropoutGPT-4Label SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

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