Papers › RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness

RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness

27 May 2024CVPR 2025 1arXiv:2405.17220archive 2025-07-28

Tianyu Yu, Haoye Zhang, Qiming Li, Qixin Xu, Yuan YAO, Da Chen, Xiaoman Lu, Ganqu Cui, Yunkai Dang, Taiwen He, Xiaocheng Feng, Jun Song, Bo Zheng, Zhiyuan Liu, Tat-Seng Chua, Maosong Sun

Traditional feedback learning for hallucination reduction relies on labor-intensive manual labeling or expensive proprietary models. This leaves the community without foundational knowledge about how to build high-quality feedback with open-source MLLMs. In this work, we introduce RLAIF-V, a novel framework that aligns MLLMs in a fully open-source paradigm. RLAIF-V maximally explores open-source MLLMs from two perspectives, including high-quality feedback data generation for preference learning and self-feedback guidance for inference-time scaling. Extensive experiments on six benchmarks in both automatic and human evaluation show that RLAIF-V substantially enhances the trustworthiness of models at both preference learning and inference time. RLAIF-V 7B reduces object hallucination by 80.7% and overall hallucination by 33.7%. Remarkably, RLAIF-V 12B further reveals the self-alignment potential of open-source MLLMs, where the model can learn from feedback of itself to achieve super GPT-4V trustworthiness.

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 22 code samples harvested from 3 repositories linked to this paper: 17 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 2 honoured, 3 violated, 2 with no contract checked; 10 where Syntology's instrument failed) · 5 unverified. Where Syntology's instrument failed: 7 where our contract draft was wrong; 3 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 15 samples from 2 repositories, 11 ran (5 with no instrument failure; 6 where Syntology's instrument failed); community (archive-listed): 7 samples from 1 repository, 6 ran (2 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.

No sample is linked to this paper from 2 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 3 of the 17 samples that ran have no recorded date.

openbmb/omnilmm officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 14 samples, 10 ran (5 instrument-failed) · 4 unverified report
rlhf-v/rlaif-v officialmentioned in papermentioned on GitHubpytorchSyntology: 1 sample, 1 ran (1 instrument-failed) · 0 unverified report
OpenBMB/MiniCPM-o mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to MiniCPM-V: A GPT-4V Level MLLM on Your Phone (arXiv:2408.01800; that paper's own run record: 11 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 2 violated, 2 with no contract checked; 6 where Syntology's instrument failed) · 3 unverified); nothing checks that they implement this paper's method. report
openbmb/minicpm-v mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to MiniCPM-V: A GPT-4V Level MLLM on Your Phone (arXiv:2408.01800; that paper's own run record: 11 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 2 violated, 2 with no contract checked; 6 where Syntology's instrument failed) · 3 unverified); nothing checks that they implement this paper's method. report
rlhf-v/rlhf-v mentioned on GitHubSyntology: 7 samples, 6 ran (4 instrument-failed) · 1 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

22 samples harvested: 17 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 2 honoured, 3 violated, 2 with no contract checked; 10 where Syntology's instrument failed) · 5 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
3ran · violated contract
7ran · our draft was wrong
3ran; the contract check could not run
2ran
5unverified

Licence: 17 of the 22 samples are pointer only, meaning Syntology does not serve that copy's text. For 9 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 3 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”. “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.

openbmb/omnilmm official repository 14 samples, 10 ran (5 instrument-failed) · 4 unverified
autocontrast_func in 17 papers omnilmm/model/utils.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) here; the stored copy is pointer only · a95e6290546a4d45 · report fixture: signature img:ndarray/3/int, cutoff:int/0/int · returned ndarray · level recorded 2026-09-22
conversation_to_ids_minicpm finetune/dataset.py ran Apache-2.0 (permissive) here; the stored copy is pointer only · d6c1484ab50b4e26 · report fixture: signature conversation:list/1/dict, tokenizer:MockTokenizer/0/? · returned tuple · level recorded 2026-09-25
data_collator in 11 papers finetune/dataset.py ran Apache-2.0 (permissive) here; the stored copy is pointer only · 4212780bf7da9ddd · report fixture: signature examples:list/1/dict, padding_value:int/0/int, max_length:int/0/int · returned dict · level recorded 2026-09-25
equalize_func in 17 papers omnilmm/model/utils.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) here; the stored copy is pointer only · eb4f4c5587a11a2c · report fixture: signature img:ndarray/3/int · returned ndarray · level recorded 2026-09-22
get_2d_sincos_pos_embed in 170 papers omnilmm/model/resampler.py ran · honoured contract fingerprinted Apache-2.0 (permissive) here; the stored copy is pointer only · 77e8a3ac46f3afec · report fixture: signature embed_dim:int/0/int, grid_size:int/0/int, cls_token:bool/0/bool · returned ndarray · level recorded 2026-09-24
get_abs_pos in 15 papers omnilmm/model/resampler.py ran · our draft was wrong fingerprinted Apache-2.0 (permissive) here; the stored copy is pointer only · 50c4398a8645438a · report fixture: signature abs_pos:Tensor/2/float, tgt_size:int/0/int · returned Tensor · level recorded 2026-09-22
get_parameter_number in 19 papers finetune/finetune.py ran · our draft was wrong Apache-2.0 (permissive) · 1a3baf406bc7de64 · report fixture: signature model:callable/0/? · returned dict · level recorded 2026-09-22
identity_func in 19 papers omnilmm/model/utils.py ran · violated contract fingerprinted Apache-2.0 (permissive) here; the stored copy is pointer only · 1dd2ab205689d45b · report fixture: signature img:Tensor/4/float · returned Tensor · level recorded 2026-09-22
upload_img web_demos/web_demo.py ran; the contract check could not run (RAISES) Apache-2.0 (permissive) · 2a1ffa5243150412 · report fixture: signature image:ndarray/3/int, _chatbot:list/1/empty, _app_session:dict/0/? · returned tuple · level recorded 2026-09-22
violates_moderation in 59 papers omnilmm/utils.py ran · violated contract Apache-2.0 (permissive) here; the stored copy is pointer only · f9939a84b9a65279 · report fixture: signature text:str/0/str · returned bool · level recorded 2026-09-26
conversation_to_ids finetune/dataset.py unverified Apache-2.0 (permissive) · a32cd6f57d6772ae · report
get_2d_sincos_pos_embed_from_grid in 140 papers omnilmm/model/resampler.py unverified Apache-2.0 (permissive) here; the stored copy is pointer only · 2884deadbb637c3e · report
omni_preprocess omnilmm/train/train_utils.py unverified Apache-2.0 (permissive) · b474c783b31ec8ac · report
pretty_print_semaphore in 55 papers omnilmm/utils.py unverified Apache-2.0 (permissive) · 37899f22fb191b37 · report
rlhf-v/rlaif-v official repository 1 sample, 1 ran (1 instrument-failed) · 0 unverified
expand_question_into_multimodal chat.py ran · our draft was wrong no licence file found · pointer only · f067ba49686d26ba · report fixture: signature question_text:list/1/dict, image_token_len:int/0/int, im_st_token:str/0/str, im_ed_token:str/0/str, im_patch_token:str/0/str · returned list · level recorded 2026-09-22
rlhf-v/rlhf-v community (archive-listed) 7 samples, 6 ran (4 instrument-failed) · 1 unverified
cal_informative eval/summarize_gpt_mmhal_review.py ran · honoured contract no licence file found · pointer only · 03a2b1b7f5fd092e · report fixture: signature path:str/0/str, return_meta:bool/0/bool · returned float64 · level recorded 2026-09-24
combine_coco_captions in 11 papers eval/eval_gpt_obj_halbench.py ran · our draft was wrong no licence file found · pointer only · 25585704d77cafb4 · report fixture: signature annotation_path:str/0/str · returned dict · level recorded, date not recorded
parse_object_list eval/eval_gpt_obj_halbench.py ran · our draft was wrong no licence file found · pointer only · 160038ef11a44314 · report fixture: signature content:str/0/str · returned list · level recorded 2026-09-24
parse_score in 49 papers eval/eval_gpt_review_llava.py ran · violated contract fingerprinted no licence file found · pointer only · 763f6fe20bd52fbb · report fixture: signature review:str/0/str · returned list · level recorded, date not recorded
preprocess_coh_results eval/eval_gpt_obj_halbench.py ran · our draft was wrong no licence file found · pointer only · f7745bc5cafc18b6 · report fixture: signature caps:list/1/dict · returned list · level recorded, date not recorded
read_json in 250 papers eval/change_mmhal_predict_template.py ran · our draft was wrong no licence file found · pointer only · f8c65c92ec213846 · report fixture: signature path:str/0/str · returned dict · level recorded 2026-09-22
get_eval in 18 papers eval/gpt4_grpc.py unverified no licence file found · pointer only · 77010a13f8bd8d11 · report

Tasks

HallucinationImage CaptioningObject HallucinationVisual Question Answering

Datasets

Introduced by this paper, per the archive.

RLAIF-V Dataset

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Captioning Object HalBench RLAIF-V 7B chair_i 4.3 #2 of 3 Archive leaderboard report
Image Captioning Object HalBench RLAIF-V 7B chair_s 8.5 #2 of 3 Archive leaderboard report
Image Captioning Object HalBench RLAIF-V 12B chair_i 1.8 #3 of 3 Archive leaderboard report
Image Captioning Object HalBench RLAIF-V 12B chair_s 3.3 #3 of 3 Archive leaderboard report
Visual Question Answering AMBER RLAIF-V 12B Accuracy 88 #1 of 1 Archive leaderboard report
Visual Question Answering AMBER RLAIF-V 12B F1 90.9 #1 of 1 Archive leaderboard report
Visual Question Answering MMHal-Bench RLAIF-V 12B Hallucination Rate 29.2 #1 of 2 Archive leaderboard report
Visual Question Answering MMHal-Bench RLAIF-V 12B Score 3.36 #1 of 2 Archive leaderboard report
Visual Question Answering MMHal-Bench RLAIF-V 7B Hallucination Rate 29.2 #2 of 2 Archive leaderboard report
Visual Question Answering MMHal-Bench RLAIF-V 7B Score 3.06 #2 of 2 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