Papers › Sigmoid Loss for Language Image Pre-Training

Sigmoid Loss for Language Image Pre-Training

27 Mar 2023ICCV 2023 1arXiv:2303.15343archive 2025-07-28

Xiaohua Zhai, Basil Mustafa, Alexander Kolesnikov, Lucas Beyer

We propose a simple pairwise Sigmoid loss for Language-Image Pre-training (SigLIP). Unlike standard contrastive learning with softmax normalization, the sigmoid loss operates solely on image-text pairs and does not require a global view of the pairwise similarities for normalization. The sigmoid loss simultaneously allows further scaling up the batch size, while also performing better at smaller batch sizes. Combined with Locked-image Tuning, with only four TPUv4 chips, we train a SigLiT model that achieves 84.5% ImageNet zero-shot accuracy in two days. The disentanglement of the batch size from the loss further allows us to study the impact of examples vs pairs and negative to positive ratio. Finally, we push the batch size to the extreme, up to one million, and find that the benefits of growing batch size quickly diminish, with a more reasonable batch size of 32k being sufficient. We release our models at https://github.com/google-research/big_vision and hope our research motivates further explorations in improving the quality and efficiency of language-image pre-training.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

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="2303.15343")

Code

Syntology Ran 14 of 29 code samples harvested from 4 repositories linked to this paper; 15 have no recorded run. Of those that ran: 3 ran · our draft was wrong; 11 ran with no contract checked.

By repository: community (archive-listed): 25 samples from 4 repositories, 12 ran; 4 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

google-research/big_vision officialmentioned in papermentioned on GitHubjax report
apple/ml-mobileclip mentioned on GitHubpytorch report
borisdayma/clip-jax mentioned on GitHubjaxApache-2.0 report
filipbasara0/relic mentioned on GitHubpytorch report
filipbasara0/simple-clip mentioned on GitHubpytorch report
huggingface/transformers mentioned on GitHubpytorch report
merveenoyan/siglip mentioned on GitHubpytorchApache-2.0 report
mlfoundations/open_clip mentioned on GitHubpytorch report
morrisfl/unifex mentioned on GitHubpytorch report
ramanakshay/clip mentioned on GitHubpytorch 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

29 samples harvested; 14 ran; 0 honoured the contract we drafted; 15 have no recorded run. Read from Syntology's graph 2026-09-25; that is when this build read the record, not when the samples ran.

3ran · our draft was wrong
11ran
15unverified

Licence: 25 of the 29 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 4 repositories linked to this paper, official or community; each sample names its own and says which. Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “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.

GlobalPool apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · metamorphic tier: invariant fingerprinted licence not identified · pointer only · af515d1162a90440 · report
GlobalPool2D apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · metamorphic tier: invariant licence not identified · pointer only · 21e0abb626a7ae62 · report
LayerNormFP32 apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · metamorphic tier: invariant fingerprinted licence not identified · pointer only · c4c619fef7d59d34 · report
LearnablePositionalEmbedding apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · metamorphic tier: well formed licence not identified · pointer only · 0748e821730b3fc1 · report
MultiHeadAttention apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · metamorphic tier: invariant fingerprinted licence not identified · pointer only · 88f8bcf4b76c2844 · report
NeighbourExchange mlfoundations/open_clip/src/open_clip/loss.py community (archive-listed) ran licence not identified · pointer only · 997398ad737e7cd2 · report
NeighbourExchangeBidir mlfoundations/open_clip/src/open_clip/loss.py community (archive-listed) ran licence not identified · pointer only · 7caa2b64e6a6f021 · report
PositionalEmbedding apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · metamorphic tier: well formed licence not identified · pointer only · 1e38402044297bef · report
SigLipLoss mlfoundations/open_clip/src/open_clip/loss.py community (archive-listed) ran licence not identified · pointer only · 2d2c2d56e476e772 · report
get_normalization_layer apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) ran · our draft was wrong licence not identified · pointer only · 59434b67e9a74429 · report
postprocess merveenoyan/siglip/clip_siglip.py community (archive-listed) ran Apache-2.0 (permissive) · 3586d486792b020d · report
postprocess_siglip merveenoyan/siglip/clip_siglip.py community (archive-listed) ran Apache-2.0 (permissive) · ad49d2b1bad09561 · report
CLIP ramanakshay/clip/src/model/model.py community (archive-listed) unverified MIT (permissive) · 7f362a66c1752126 · report
CLIP apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · a8fa6ab303249a21 · report
CLIPModel ramanakshay/clip/src/model/model.py community (archive-listed) unverified MIT (permissive) · e80ed1f8e5c3f5a8 · report
ConvFFN apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · df9eaf41a5c09a07 · report
MCi apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · 579a47ae5bfe8248 · report
MobileOneBlock apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · 695f06cd58c0597e · report
RepMixer apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · 43cafd406a2da3f3 · report
RepMixerBlock apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · efcf8d9368ee9d08 · report
SEBlock apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · 49fac99e1f1a0270 · report
TextTransformer apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · 88fa799b7ff64ccb · report
TransformerEncoder apple/ml-mobileclip/mobileclip/clip.py community (archive-listed) unverified licence not identified · pointer only · 35094b7d325013c5 · report
neighbour_exchange_bidir_with_grad mlfoundations/open_clip/src/open_clip/loss.py community (archive-listed) unverified licence not identified · pointer only · 4a8006bab44bb658 · report
neighbour_exchange_with_grad mlfoundations/open_clip/src/open_clip/loss.py community (archive-listed) unverified licence not identified · pointer only · 12c8728ab37fdce8 · report
relic_loss identical code first harvested elsewhere ran · our draft was wrong licence of this copy not recorded · eb62100582261eb4 · report
siglip_loss identical code first harvested elsewhere ran · our draft was wrong fingerprinted licence of this copy not recorded · 67c4ef5004396e02 · report
neighbour_exchange identical code first harvested elsewhere unverified licence of this copy not recorded · e332856e3c2fc814 · report
neighbour_exchange_bidir identical code first harvested elsewhere unverified licence of this copy not recorded · 5b1fd364afcf3c05 · report

Tasks

Contrastive LearningDisentanglementImage-to-Text RetrievalZero-shot Text-to-Image Retrieval

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image-to-Text Retrieval COCO SigLIP (ViT-L, zero-shot) Recall@1 70.6 #1 of 1 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

BASECLIPContrastive LearningSoftmax

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