SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.02848 · 2020

Unsupervised pretraining transfers well across languages

arXiv · PDF · Open in the Atlas

Code that ran

2 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 1 no recorded run), measured 2026-11-21; 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/2002.02848.json · badge: /api/ran/2002.02848.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
facebookresearch/CPC_audio canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
FunctionStatusWhere it lives
parse_args Ran
ran; our contract draft was wrong, not the code
facebookresearch/CPC_audio/cpc/eval/linear_separability.py
code served (permissive licence) · get_code("f5d2474d463058b0")
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 Namespace. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 pathlib import Path
import tempfile
import os

torch.manual_seed(0)

# Create temporary directories and files for the function to use
temp_dir = tempfile.mkdtemp()
db_dir = os.path.join(temp_dir, "audio_db")
os.makedirs(db_dir, exist_ok=True)

train_list_path = os.path.join(temp_dir, "train_list.txt")
val_list_path = os.path.join(temp_dir, "val_list.txt")
phone_labels_path = os.path.join(temp_dir, "phone_labels.txt")
checkpoint_dir = os.path.join(temp_dir, "checkpoints")
os.makedirs(checkpoint_dir, exist_ok=True)

# Create dummy files
with open(train_list_path, 'w') as f:
    f.write("sample1\nsample2\n")

with open(val_list_path, 'w') as f:
    f.write("sample3\nsample4\n")

with open(phone_labels_path, 'w') as f:
    f.write("phone_data\n")

# Create a dummy checkpoint file
checkpoint_path = os.path.join(temp_dir, "model.pt")
torch.save({"state_dict": {}}, checkpoint_path)

# Construct argv list with all required and optional arguments
argv = [
    db_dir,                          # pathDB
    train_list_path,                 # pathTrain
    val_list_path,                   # pathVal
    checkpoint_path,                 # load (positional, nargs='*')
    "--pathPhone", phone_labels_path,
    "--CTC",
    "--pathCheckpoint", checkpoint_dir,
    "--nGPU", "1",
    "--batchSizeGPU", "8",
    "--n_epoch", "10",
    "--debug",
    "--unfrozen",
    "--no_pretraining",
    "--file_extension", ".flac",
    "--save_step", "5",
    "--get_encoded",
    "--lr", "0.0002",
    "--beta1", "0.9",
    "--beta2", "0.999",
    "--epsilon", "2e-08",
    "--ignore_cache",
    "--size_window", "20480",
]

ARGS = {
    "argv": argv,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Namespace — a Namespace output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Namespace 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
reduce_sparse_data Ran
ran, honoured the contract we drafted
facebookresearch/CPC_audio/cpc/eval/ABX.py
code served (permissive licence) · get_code("b4c2b5de7a1d9b50")
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 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
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# - quotient: used in division operation (/)
# - divisor: used in comparison (==), multiplication (*), and division (/)
# Both should be tensors or arrays that support these operations

quotient = torch.randn(2, 4)
divisor = torch.randn(2, 4)

ARGS = {
    "quotient": quotient,
    "divisor": divisor,
}
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 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
parse_args Not yet run
unverified (harvested, no recorded run)
facebookresearch/CPC_audio/cpc/eval/ABX.py
code served (permissive licence) · get_code("9891603bc50b0600")
How it ran

no invocation 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, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 1 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
parse_args 2 other papers
2 later
2104.11946 (2021-04)
Aligned Contrastive Predictive Coding
2110.15909 (2021-10)
Contrastive prediction strategies for unsupervised segmentation and categorization of phonemes and words
reduce_sparse_data 2 other papers
2 later
2104.11946 (2021-04)
Aligned Contrastive Predictive Coding
2110.15909 (2021-10)
Contrastive prediction strategies for unsupervised segmentation and categorization of phonemes and words

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 2026-11-21.

Repositories linked to this paper

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

Abstract

Cross-lingual and multi-lingual training of Automatic Speech Recognition (ASR) has been extensively investigated in the supervised setting. This assumes the existence of a parallel corpus of speech and orthographic transcriptions. Recently, contrastive predictive coding (CPC) algorithms have been proposed to pretrain ASR systems with unlabelled data. In this work, we investigate whether unsupervised pretraining transfers well across languages. We show that a slight modification of the CPC pretraining extracts features that transfer well to other languages, being on par or even outperforming supervised pretraining. This shows the potential of unsupervised methods for languages with few linguistic resources.

For agents

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

get_harvested_code_for_paper("2002.02848")
get_code_for_paper("2002.02848")
have("2002.02848")

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

curl https://syntology.ai/api/ran/2002.02848.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.02848.svg)](https://syntology.ai/paper/2002.02848)

Connect an agent — have() is free.