SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2105.02412 · 2021

Handwritten Mathematical Expression Recognition with Bidirectionally Trained Transformer

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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; 3 failures of our instrument; 0 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/2105.02412.json · badge: /api/ran/2105.02412.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
Green-Wood/BTTR canonical (classifier label)
classifier basis: census
3 ran · 0 no recorded run
FunctionStatusWhere it lives
data_iterator Ran
ran; our contract draft was wrong, not the code
Green-Wood/BTTR/bttr/datamodule/datamodule.py
code served (permissive licence) · get_code("c2c5353ea8549a2a")
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 list. 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
from PIL import Image
import numpy as np

torch.manual_seed(0)

# Create sample images
img1 = Image.new('RGB', (100, 80))  # 8000 pixels
img2 = Image.new('RGB', (120, 90))  # 10800 pixels

# Create Data: List[Tuple[str, Image.Image, List[str]]]
data = [
    ('image1.jpg', img1, ['hello', 'world']),
    ('image2.jpg', img2, ['foo', 'bar', 'baz']),
]

ARGS = {
    'data': data,
    'batch_size': 2,
    'batch_Imagesize': 320000.0,
    'maxlen': 200,
    'maxImagesize': 320000.0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is 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
extract_data Ran
ran; our contract draft was wrong, not the code
Green-Wood/BTTR/bttr/datamodule/datamodule.py
code served (permissive licence) · get_code("c5115025a3ddb95b")
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 list. 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 io
import torch
import numpy as np
from zipfile import ZipFile
from PIL import Image

# Create an in-memory zip file with test data
zip_buffer = io.BytesIO()
with ZipFile(zip_buffer, 'w') as zf:
    # Create a caption.txt file with test data
    caption_content = b"img001 the quick brown fox\nimg002 a simple test image\n"
    zf.writestr('test_dir/caption.txt', caption_content)
    
    # Create test BMP images
    for i in range(1, 3):
        img = Image.new('RGB', (10, 10), color=(73 + i, 109 + i, 137 + i))
        img_buffer = io.BytesIO()
        img.save(img_buffer, format='BMP')
        img_buffer.seek(0)
        zf.writestr(f'test_dir/img00{i}.bmp', img_buffer.read())

# Reset buffer position for reading
zip_buffer.seek(0)

# Open the archive
archive = ZipFile(zip_buffer, 'r')

ARGS = {
    'archive': archive,
    'dir_name': 'test_dir'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is 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
rotate_every_two Ran
ran; our contract draft was wrong, not the code
Green-Wood/BTTR/bttr/model/pos_enc.py
pointer only (licence: MIT); the stored copy get_code serves, from qingzhenduyu/tamer, is pointer only (licence: NONE) · get_code("e06d2f5f2824ecc6")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from qingzhenduyu/tamer, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 2 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
rotate_every_two 2 other papers
2 later
2407.07764 (2024-07)
PosFormer: Recognizing Complex Handwritten Mathematical Expression with Position Forest Transformer
2408.08578 (2024-08)
TAMER: Tree-Aware Transformer for Handwritten Mathematical Expression Recognition

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

Encoder-decoder models have made great progress on handwritten mathematical expression recognition recently. However, it is still a challenge for existing methods to assign attention to image features accurately. Moreover, those encoder-decoder models usually adopt RNN-based models in their decoder part, which makes them inefficient in processing long $\LaTeX{}$ sequences. In this paper, a transformer-based decoder is employed to replace RNN-based ones, which makes the whole model architecture very concise. Furthermore, a novel training strategy is introduced to fully exploit the potential of the transformer in bidirectional language modeling. Compared to several methods that do not use data augmentation, experiments demonstrate that our model improves the ExpRate of current state-of-the-art methods on CROHME 2014 by 2.23%. Similarly, on CROHME 2016 and CROHME 2019, we improve the ExpRate by 1.92% and 2.28% respectively.

For agents

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

get_harvested_code_for_paper("2105.02412")
get_code_for_paper("2105.02412")
have("2105.02412")

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

curl https://syntology.ai/api/ran/2105.02412.json

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

[![Syntology run record](https://syntology.ai/api/ran/2105.02412.svg)](https://syntology.ai/paper/2105.02412)

Connect an agent — have() is free.