SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2202.10401 · CVPR · 2022

Vision-Language Pre-Training with Triple Contrastive Learning

Junzhou Huang, Jinyu Yang, Yi Xu, Belinda Zeng, Trishul Chilimbi, Son Tran, Liqun Chen, Jiali Duan, Sampath Chanda

arXiv · PDF · Open in the Atlas

Code that ran

1 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 1 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2202.10401.json · badge: /api/ran/2202.10401.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
uta-smile/TCL canonical (classifier label)
classifier basis not recorded on the link
1 ran · 1 no recorded run
FunctionStatusWhere it lives
itm_eval Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
uta-smile/TCL/Retrieval.py
code served (permissive licence) · get_code("c7f85279d772ea19")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import torch

torch.manual_seed(0)

# scores_i2t: shape (num_images, num_texts) - similarity scores from images to texts
# Each row is scores for one image against all texts
scores_i2t = np.random.rand(2, 4).astype(np.float32)

# scores_t2i: shape (num_texts, num_images) - similarity scores from texts to images
# Each row is scores for one text against all images
scores_t2i = np.random.rand(4, 2).astype(np.float32)

# txt2img: dict mapping text index to image index
# For each text, which image is the correct match
txt2img = {
    0: 0,
    1: 1,
    2: 0,
    3: 1
}

# img2txt: dict mapping image index to list of text indices
# For each image, which texts are valid matches
img2txt = {
    0: [0, 2],
    1: [1, 3]
}

ARGS = {
    'scores_i2t': scores_i2t,
    'scores_t2i': scores_t2i,
    'txt2img': txt2img,
    'img2txt': img2txt
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
concat_all_gather Not yet run
unverified (harvested, no recorded run)
uta-smile/TCL/models/model_pretrain.py
code served (permissive licence) · get_code("73cecca9f3575f09")
How it ran

no invocation recorded.

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 1 the same month, 2 later) and 0 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
concat_all_gather 138 other papers
36 dated earlier, 3 the same month, 97 later, 2 same year, order not determinable · 33 only as an assembly brick
1710.09412 (2017-10)
mixup: BEYOND EMPIRICAL RISK MINIMIZATION
1905.04899 (2019-05)
CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features
1905.12794 (2019-05)
Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback
and 135 more not named here
itm_eval 13 other papers
1 dated earlier, 12 later
2107.07651 (2021-07)
Align before Fuse: Vision and Language Representation Learning with Momentum Distillation
2206.00621 (2022-06)
Cross-View Language Modeling: Towards Unified Cross-Lingual Cross-Modal Pre-training
2211.12402 (2022-11)
X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks
and 10 more not named here

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Vision-language representation learning largely benefits from image-text alignment through contrastive losses (e.g., InfoNCE loss). The success of this alignment strategy is attributed to its capability in maximizing the mutual information (MI) between an image and its matched text. However, simply performing cross-modal alignment (CMA) ignores data potential within each modality, which may result in degraded representations. For instance, although CMA-based models are able to map image-text pairs close together in the embedding space, they fail to ensure that similar inputs from the same modality stay close by. This problem can get even worse when the pre-training data is noisy. In this paper, we propose triple contrastive learning (TCL) for visionlanguage pre-training by leveraging both cross-modal and intra-modal self-supervision. Besides CMA, TCL introduces an intra-modal contrastive objective to provide complementary benefits in representation learning. To take advantage of localized and structural information from image and text input, TCL further maximizes the average MI between local regions of image/text and their global summary. To the best of our knowledge, ours is the first work that takes into account local structure information for multi-modality representation learning. Experimental evaluations show that our approach is competitive and achieves the new state of the art on various common down-stream vision-language tasks such as image-text retrieval and visual question answering.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2202.10401")
get_code_for_paper("2202.10401")
have("2202.10401")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2202.10401.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2202.10401.svg)](https://syntology.ai/paper/2202.10401)

Connect an agent — have() is free.