SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2211.00262 · 2022

Training Vision-Language Models with Less Bimodal Supervision

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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 3 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2211.00262.json · badge: /api/ran/2211.00262.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
eladsegal/less-bimodal-sup canonical (classifier label)
classifier basis not recorded on the link
3 ran · 1 no recorded run
FunctionStatusWhere it lives
identity_func Ran
ran on a synthesized input
eladsegal/less-bimodal-sup/src/evaluators/hf_evaluator.py
code served (permissive licence) · get_code("d715edbedc06dfd2")
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, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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 torch

torch.manual_seed(0)

ARGS = {
    "id_": "test_id_001",
    "x": torch.randn(2, 4)
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
unflatten Ran
ran on a synthesized input
eladsegal/less-bimodal-sup/utils/jsonnet.py
code served (permissive licence) · get_code("225dafec388f7a2f")
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, through the batch API) 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 torch
import numpy as np
from typing import Any, Dict

# Construct a valid flat_dict argument
flat_dict = {
    "a.b": 0,
    "a.c": 1,
    "d": 2,
    "e.f.g": 3,
    "e.f.h": 4,
}

ARGS = {
    "flat_dict": flat_dict,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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
with_fallback Ran
ran on a synthesized input
eladsegal/less-bimodal-sup/utils/jsonnet.py
code served (permissive licence) · get_code("d467682ffde173a4")
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, through the batch API) 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 copy
from typing import Any, Dict

# Test case 1: Simple dict merge with non-overlapping keys
preferred_1 = {"a": 1, "b": 2}
fallback_1 = {"c": 3, "d": 4}

# Test case 2: Dict merge with overlapping keys (scalars)
preferred_2 = {"x": 10, "y": 20}
fallback_2 = {"x": 100, "y": 200, "z": 300}

# Test case 3: Nested dict merge
preferred_3 = {"config": {"learning_rate": 0.001, "batch_size": 32}}
fallback_3 = {"config": {"learning_rate": 0.01, "epochs": 100}, "model": "bert"}

# Test case 4: Dict with list values (list concatenation)
preferred_4 = {"layers": [5, 6]}
fallback_4 = {"layers": [1, 2, 3], "other": "value"}

# Test case 5: Dict with mixed nested structures
preferred_5 = {
    "optimizer": {"params": {"weight_decay": 1e-5}},
    "training": [100, 200]
}
fallback_5 = {
    "optimizer": {"params": {"lr": 0.001, "weight_decay": 1e-4}},
    "training": [10, 20, 30],
    "validation": True
}

# Use the simplest case for ARGS
ARGS = {
    "preferred": preferred_1,
    "fallback": fallback_1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — 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
force_eval Not yet run
unverified (harvested, no recorded run)
eladsegal/less-bimodal-sup/src/callbacks/freezer.py
code served (permissive licence) · get_code("cb025115712dfc97")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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
unflatten 1 other paper
1 dated earlier
2201.03533 (2022-01)
SCROLLS: Standardized CompaRison Over Long Language Sequences

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

Standard practice in pretraining multimodal models, such as vision-language models, is to rely on pairs of aligned inputs from both modalities, for example, aligned image-text pairs. However, such pairs can be difficult to obtain in low-resource settings and for some modality pairs (e.g., structured tables and images). In this work, we investigate the extent to which we can reduce the reliance on such parallel data, which we term \emph{bimodal supervision}, and use models that are pretrained on each modality independently. We experiment with a high-performing vision-language model, and analyze the effect of bimodal supervision on three vision-language tasks. We find that on simpler tasks, such as VQAv2 and GQA, one can eliminate bimodal supervision completely, suffering only a minor loss in performance. Conversely, for NLVR2, which requires more complex reasoning, training without bimodal supervision leads to random performance. Nevertheless, using only 5\% of the bimodal data (142K images along with their captions), or leveraging weak supervision in the form of a list of machine-generated labels for each image, leads to only a moderate degradation compared to using 3M image-text pairs: 74\%$\rightarrow$$\sim$70\%. Our code is available at https://github.com/eladsegal/less-bimodal-sup.

For agents

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

get_harvested_code_for_paper("2211.00262")
get_code_for_paper("2211.00262")
have("2211.00262")

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

curl https://syntology.ai/api/ran/2211.00262.json

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

[![Syntology run record](https://syntology.ai/api/ran/2211.00262.svg)](https://syntology.ai/paper/2211.00262)

Connect an agent — have() is free.