SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2607.00890 · 2026

MultiSynt/MT: Trillion-Token Multi-Parallel Pre-Training Data Translated Across 36 Languages

André Martins, David Salinas, Jenia Jitsev, Zihao Li, Sampo Pyysalo, Jan Hajič, Jenna Kanerva, Jörg Tiedemann, Filip Ginter, Andrey Kutuzov, Matthias Lindemann, Gema Ramírez-Sánchez, and 10 more

arXiv · PDF

Code that ran

4 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 0 have no contract verdict; 4 failures of our instrument; 0 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/2607.00890.json · badge: /api/ran/2607.00890.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
Helsinki-NLP/Opus-MT canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
FunctionStatusWhere it lives
create_parser Ran
ran; our contract draft was wrong, not the code
Helsinki-NLP/Opus-MT/apply_bpe.py
code served (permissive licence) · get_code("f49716cdf8379c4f")
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 ArgumentParser. 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 sys
import argparse
import tempfile
import io

# Create temporary files for the required file arguments
codes_content = """#version: 0.2
l o
t h
i s
"""
codes_file = io.StringIO(codes_content)

input_content = """hello world
this is a test
"""
input_file = io.StringIO(input_content)

output_file = io.StringIO()

vocabulary_content = """hello 100
world 95
this 90
is 85
a 80
test 75
"""
vocabulary_file = io.StringIO(vocabulary_content)

ARGS = {
    'subparsers': None,
}
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 ArgumentParser — a ArgumentParser output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ArgumentParser 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
encode Ran
ran; our contract draft was wrong, not the code
Helsinki-NLP/Opus-MT/apply_bpe.py
code served (permissive licence) · get_code("d4c1f9f4c9e3d8f8")
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, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 re

# Build minimal arguments for the encode function
# The function applies BPE (Byte Pair Encoding) merges to encode a word

# Small BPE codes dictionary mapping bigram pairs to merge indices
bpe_codes = {
    ('l', 'o'): 0,
    ('lo', 'w'): 1,
    ('e', 'r'): 2,
}

# Reverse BPE codes for recursive_split (maps merged form back to components)
bpe_codes_reverse = {
    'lo</w>': ('l', 'o</w>'),
    'low</w>': ('lo', 'w</w>'),
    'er</w>': ('e', 'r</w>'),
}

# Vocabulary - set of valid tokens
vocab = {
    'l', 'o', 'w', 'e', 'r',
    'lo', 'low', 'er',
    'l</w>', 'o</w>', 'w</w>', 'e</w>', 'r</w>',
    'lo</w>', 'low</w>', 'er</w>',
}

# Cache for encoded words
cache = {}

# Glossaries - regex pattern for special words to not encode
glossaries = ['<unk>', '<pad>']

# Test with version (0, 2) which is the more common one
orig = 'lower'
separator = '</w>'
version = (0, 2)

ARGS = {
    'orig': orig,
    'bpe_codes': bpe_codes,
    'bpe_codes_reverse': bpe_codes_reverse,
    'vocab': vocab,
    'separator': separator,
    'version': version,
    'cache': cache,
    'glossaries': glossaries,
}
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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
get_pairs Ran
ran; our contract draft was wrong, not the code
Helsinki-NLP/Opus-MT/apply_bpe.py
code served (permissive licence) · get_code("290febe7a42e8479")
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 set. 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

# The function get_pairs takes a word parameter
# word should be a tuple of symbols (variable-length strings)
# The function iterates through the tuple and creates pairs of consecutive elements

# Create a simple word as a tuple of symbols
word = ("h", "e", "l", "l", "o")

ARGS = {
    "word": word
}
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 set — a set output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a set 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
ms Ran
ran; our contract draft was wrong, not the code
Helsinki-NLP/Opus-MT/elg/generate_metadata.py
code served (permissive licence) · get_code("21d537bb993fcd9d")
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 str. 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.

ms_namespace_uri = 'http://w3id.org/meta-share/meta-share/'
ms_prefix = '{%s}' % ms_namespace_uri

def ms(tag):
    return ms_prefix + tag

ARGS = {
    'tag': 'element'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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

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, 1 the same month, 1 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
get_pairs 258 other papers
254 dated earlier, 2 the same month, 1 later, 1 undated · 4 only as an assembly brick
1802.06474 (2018-02)
A Closed-form Solution to Photorealistic Image Stylization
1803.01128 (2018-03)
Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial Examples
1809.00125 (2018-09)
Simple Fusion: Return of the Language Model
and 255 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

Open web-scale pre-training corpora remain concentrated in English, limiting multilingual LLM development. We introduce Mul-tiSynt/MT, an open synthetic parallel corpus with approximately 4.8 trillion target-language tokens across 36 languages, produced by translating 100 billion high-quality Nemotron-CC tokens with TOWER+ and OPUS-MT/HPLT-MT systems. For many medium-and lowerresource European languages, this is the largest openly available pre-training resource. On a broad multilingual benchmark suite, reference LLMs trained on MultiSynt/MT reach the final score of HPLT 2.0, a native-data baseline, using roughly 72% fewer pre-training tokens, and outperform it by approximately 15% relative at a matched 100B-token training budget. Our analyses also identify evaluation blind spots: standard multiple-choice benchmarks miss translation-quality differences that a fluency-sensitive LLM-as-judge evaluation cleanly recovers on the trained LLMs (with no fluency deficit in MultiSynt itself), and Norwegian idiomatic and culturally grounded tasks, for example, remain better served by native data. We release the corpus, including rowaligned translations from multiple systems, to support controlled research on multilingual pretraining data and evaluation.

For agents

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

get_harvested_code_for_paper("2607.00890")
get_code_for_paper("2607.00890")
have("2607.00890")

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

curl https://syntology.ai/api/ran/2607.00890.json

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

[![Syntology run record](https://syntology.ai/api/ran/2607.00890.svg)](https://syntology.ai/paper/2607.00890)

Connect an agent — have() is free.