Papers › Uncertainty-Aware Evaluation for Vision-Language Models

Uncertainty-Aware Evaluation for Vision-Language Models

22 Feb 2024arXiv:2402.14418archive 2025-07-28

Vasily Kostumov, Bulat Nutfullin, Oleg Pilipenko, Eugene Ilyushin

Vision-Language Models like GPT-4, LLaVA, and CogVLM have surged in popularity recently due to their impressive performance in several vision-language tasks. Current evaluation methods, however, overlook an essential component: uncertainty, which is crucial for a comprehensive assessment of VLMs. Addressing this oversight, we present a benchmark incorporating uncertainty quantification into evaluating VLMs. Our analysis spans 20+ VLMs, focusing on the multiple-choice Visual Question Answering (VQA) task. We examine models on 5 datasets that evaluate various vision-language capabilities. Using conformal prediction as an uncertainty estimation approach, we demonstrate that the models' uncertainty is not aligned with their accuracy. Specifically, we show that models with the highest accuracy may also have the highest uncertainty, which confirms the importance of measuring it for VLMs. Our empirical findings also reveal a correlation between model uncertainty and its language model part.

PaperPDFCodeCode Syntology ran

In Syntology 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="2402.14418")

Code

Syntology Ran 7 of 17 code samples harvested from 1 repository linked to this paper; 10 have no recorded run. Of those that ran: 1 ran · violated contract; 6 ran with no contract checked.

By repository: official repository: 17 samples from 1 repository, 7 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

ensec-ai/vlm-uncertainty-bench officialmentioned in paperpytorchMIT 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

17 samples harvested; 7 ran; 0 honoured the contract we drafted; 10 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · violated contract
6ran
10unverified

Licence: 0 of the 17 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 ensec-ai/vlm-uncertainty-bench. “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.

softmax ensec-ai/vlm-uncertainty-bench/uncertainty_quantification_via_cp.py official repository ran · violated contract fingerprinted MIT (permissive) · 7f9b15275ca5759c · report
get_options ensec-ai/vlm-uncertainty-bench/data_utils/common_utils.py official repository ran MIT (permissive) · 7ec8353674842469 · report
is_none ensec-ai/vlm-uncertainty-bench/data_utils/common_utils.py official repository ran MIT (permissive) · 25593cfba3f7b483 · report
open_json ensec-ai/vlm-uncertainty-bench/data_utils/common_utils.py official repository ran MIT (permissive) · 8a064d582b5c57ea · report
process_ai2d ensec-ai/vlm-uncertainty-bench/data_utils/ai2d.py official repository ran MIT (permissive) · ca40ba6d177df427 · report
process_seedbench ensec-ai/vlm-uncertainty-bench/data_utils/seedbench.py official repository ran MIT (permissive) · a1bac8b55fe3309a · report
process_single_sample ensec-ai/vlm-uncertainty-bench/data_utils/mmmu.py official repository ran MIT (permissive) · c1b3610ee2583172 · report
download_ai2d ensec-ai/vlm-uncertainty-bench/data_utils/ai2d.py official repository unverified MIT (permissive) · 1267ef7cf3135661 · report
download_mmbench ensec-ai/vlm-uncertainty-bench/data_utils/mmbench.py official repository unverified MIT (permissive) · a268ca9c33318e5c · report
download_seedbench ensec-ai/vlm-uncertainty-bench/data_utils/seedbench.py official repository unverified MIT (permissive) · 25a29dad236f0992 · report
get_ce ensec-ai/vlm-uncertainty-bench/uncertainty_quantification_via_cp.py official repository unverified MIT (permissive) · d40927dd38eca095 · report
get_dataset ensec-ai/vlm-uncertainty-bench/data_utils/utils.py official repository unverified MIT (permissive) · 9e203cb8767712bd · report
get_logits ensec-ai/vlm-uncertainty-bench/models_utils/utils.py official repository unverified MIT (permissive) · 4335b7a5c5a94aaf · report
parse_img_path ensec-ai/vlm-uncertainty-bench/data_utils/mmmu.py official repository unverified MIT (permissive) · de4427887bed9253 · report
process_mmmu ensec-ai/vlm-uncertainty-bench/data_utils/mmmu.py official repository unverified MIT (permissive) · eba2dfc02d4349ea · report
process_oodcv ensec-ai/vlm-uncertainty-bench/data_utils/oodcv.py official repository unverified MIT (permissive) · 5c69a8ac5c608ef9 · report
process_scienceqa ensec-ai/vlm-uncertainty-bench/data_utils/scienceqa.py official repository unverified MIT (permissive) · d2a305808682d0b6 · report

Tasks

Conformal PredictionLanguage ModelingLanguage ModellingMultiple-choiceQuestion AnsweringUncertainty QuantificationVisual Question AnsweringVisual Question Answering (VQA)

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Absolute Position EncodingsAdamAttentionBPEDense 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