SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2003.11080 · ICML · 2020

XTREME: A Massively Multilingual Multi-task Benchmark for Evaluating Cross-lingual Generalization

Graham Neubig, Junjie Hu, Sebastian Ruder, Aditya Siddhant, Orhan Firat, Melvin Johnson

arXiv · PDF · Open in the Atlas

Code that ran

7 of 10 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 2 have no contract verdict; 5 failures of our instrument; 3 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/2003.11080.json · badge: /api/ran/2003.11080.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
google-research/xtreme canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
JunjieHu/xtreme-dev canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
KyGao/token-level-classification community (classifier label)
classifier basis not recorded on the link
2 ran · 2 no recorded run
FunctionStatusWhere it lives
read_label Ran
ran; our contract draft was wrong, not the code
google-research/xtreme/evaluate.py
code served (permissive licence) · get_code("c6cdf941eeaf166a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 tempfile
import os

# Create a temporary file with label data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("label1\n")
temp_file.write("label2\n")
temp_file.write("label3\n")
temp_file.close()

ARGS = {
    'file': temp_file.name
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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
read_label Ran
ran; our contract draft was wrong, not the code
JunjieHu/xtreme-dev/evaluate.py
code served (permissive licence) · get_code("25ec008fc6924728")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 tempfile
import os

# Create a temporary file with label data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("label1\n")
temp_file.write("label2  \n")
temp_file.write("  label3\n")
temp_file.write("label4")
temp_file.close()

ARGS = {
    'file': temp_file.name
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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
read_squad Ran
ran; our contract draft was wrong, not the code
google-research/xtreme/evaluate.py
code served (permissive licence) · get_code("e78f925433eb7df1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 json
import tempfile
import os

# Create a temporary file with SQuAD-like JSON data
squad_data = {
    "data": [
        {
            "title": "Sample_Article",
            "paragraphs": [
                {
                    "context": "This is a sample context for testing.",
                    "qas": [
                        {
                            "question": "What is this?",
                            "id": "q1",
                            "answers": [
                                {
                                    "text": "sample context",
                                    "answer_start": 10
                                }
                            ]
                        }
                    ]
                }
            ]
        },
        {
            "title": "Another_Article",
            "paragraphs": [
                {
                    "context": "Another context here.",
                    "qas": [
                        {
                            "question": "Where is it?",
                            "id": "q2",
                            "answers": [
                                {
                                    "text": "here",
                                    "answer_start": 16
                                }
                            ]
                        }
                    ]
                }
            ]
        }
    ]
}

# Create a temporary file
temp_file = tempfile.NamedTemporaryFile(mode='w', suffix='.json', delete=False)
json.dump(squad_data, temp_file)
temp_file.close()

ARGS = {
    "file": temp_file.name
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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
read_squad Ran
ran on a synthesized input
KyGao/token-level-classification/evaluate.py
code served (permissive licence) · get_code("81e696f664b1ab38")
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 json
import tempfile
import os

# Create a temporary JSON file with SQuAD format data
temp_dir = tempfile.mkdtemp()
squad_file = os.path.join(temp_dir, "squad.json")

# Create a valid SQuAD v1.1 format dataset
squad_data = {
    "version": "1.1",
    "data": [
        {
            "title": "Test Article",
            "paragraphs": [
                {
                    "qas": [
                        {
                            "question": "What is the capital?",
                            "id": "q1",
                            "answers": [
                                {
                                    "text": "Paris",
                                    "answer_start": 0
                                }
                            ],
                            "is_impossible": False
                        }
                    ],
                    "context": "Paris is the capital of France."
                }
            ]
        }
    ]
}

with open(squad_file, 'w') as f:
    json.dump(squad_data, f)

ARGS = {
    "file": squad_file
}
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
read_tag Ran
ran; our contract draft was wrong, not the code
google-research/xtreme/evaluate.py
code served (permissive licence) · get_code("529118a29ca561ca")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 tempfile
import os

# Create a temporary file with NER/POS tag data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("B-PER\n")
temp_file.write("I-PER\n")
temp_file.write("O\n")
temp_file.write("\n")
temp_file.write("B-LOC\n")
temp_file.write("O\n")
temp_file.write("\n")
temp_file.write("B-ORG\n")
temp_file.close()

ARGS = {
    'file': temp_file.name
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 — 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
read_tag Ran
ran; our contract draft was wrong, not the code
JunjieHu/xtreme-dev/evaluate.py
code served (permissive licence) · get_code("eb2fa4c94377fbc5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 tempfile
import os

# Create a temporary file with sample tag data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("word1\tB-PER\n")
temp_file.write("word2\tI-PER\n")
temp_file.write("word3\tO\n")
temp_file.write("\n")
temp_file.write("word4\tB-LOC\n")
temp_file.write("word5\tI-LOC\n")
temp_file.write("\n")
temp_file.write("word6\tO\n")
temp_file.write("\n")
temp_file.close()

ARGS = {
    'file': temp_file.name
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 — 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
read_tag Ran
ran on a synthesized input
KyGao/token-level-classification/evaluate.py
code served (permissive licence) · get_code("58e41ef6e18723ab")
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 tempfile
import os

# Create a temporary file with sample data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("line1\n")
temp_file.write("line2\n")
temp_file.write("line3\n")
temp_file.write("\n")
temp_file.write("line4\n")
temp_file.write("line5\n")
temp_file.write("\n")
temp_file.close()

ARGS = {
    'file': temp_file.name
}
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 — 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
panx_tokenize_preprocess Not yet run
unverified (harvested, no recorded run)
KyGao/token-level-classification/utils_preprocess.py
code served (permissive licence) · get_code("e0964a7ee733b4f1")
How it ran

no invocation recorded.

read_squad Not yet run
unverified (harvested, no recorded run)
JunjieHu/xtreme-dev/evaluate.py
code served (permissive licence) · get_code("4b383ef4d40ea284")
How it ran

no invocation recorded.

udpos_preprocess Not yet run
unverified (harvested, no recorded run)
KyGao/token-level-classification/utils_preprocess.py
code served (permissive licence) · get_code("09344d05d53bbba7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 9 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.

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

Much recent progress in applications of machine learning models to NLP has been driven by benchmarks that evaluate models across a wide variety of tasks. However, these broad-coverage benchmarks have been mostly limited to English, and despite an increasing interest in multilingual models, a benchmark that enables the comprehensive evaluation of such methods on a diverse range of languages and tasks is still missing. To this end, we introduce the Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark, a multi-task benchmark for evaluating the cross-lingual generalization capabilities of multilingual representations across 40 languages and 9 tasks. We demonstrate that while models tested on English reach human performance on many tasks, there is still a sizable gap in the performance of cross-lingually transferred models, particularly on syntactic and sentence retrieval tasks. There is also a wide spread of results across languages. We release the benchmark 1 to encourage research on cross-lingual learning methods that transfer linguistic knowledge across a diverse and representative set of languages and tasks.

For agents

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

get_harvested_code_for_paper("2003.11080")
get_code_for_paper("2003.11080")
have("2003.11080")

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

curl https://syntology.ai/api/ran/2003.11080.json

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

[![Syntology run record](https://syntology.ai/api/ran/2003.11080.svg)](https://syntology.ai/paper/2003.11080)

Connect an agent — have() is free.