SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.07584 · ACL · 2024

UltraEval: A Lightweight Platform for Flexible and Comprehensive Evaluation for LLMs

Maosong Sun, Zhiyuan Liu, Jie Zhou, Xu Han, Jiajie Zhang, Shengding Hu, Chaoqun He, Renjie Luo, Yuanqian Zhao, Hanghao Wu

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; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 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/2404.07584.json · badge: /api/ran/2404.07584.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
openbmb/ultraeval canonical (classifier label)
classifier basis: census
2 ran · 1 no recorded run
FunctionStatusWhere it lives
get_task_path Ran
ran; our contract draft was wrong, not the code
openbmb/ultraeval/configs/make_config.py
code served (permissive licence) · get_code("747d4c40deacb4b6")
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 os
import tempfile
import shutil

# Create a temporary directory structure for testing
temp_base = tempfile.mkdtemp()

# Create dataset directories with config subdirectories and JSON files
dataset1_dir = os.path.join(temp_base, 'dataset1')
os.makedirs(dataset1_dir)
config1_dir = os.path.join(dataset1_dir, 'config')
os.makedirs(config1_dir)

# Create config files with names that will be parsed
with open(os.path.join(config1_dir, 'task_gen.json'), 'w') as f:
    f.write('{}')

with open(os.path.join(config1_dir, 'task_ppl.json'), 'w') as f:
    f.write('{}')

dataset2_dir = os.path.join(temp_base, 'dataset2')
os.makedirs(dataset2_dir)
config2_dir = os.path.join(dataset2_dir, 'config')
os.makedirs(config2_dir)

with open(os.path.join(config2_dir, 'other-task.json'), 'w') as f:
    f.write('{}')

# Define arguments for the function
ARGS = {
    'base_dir': temp_base
}
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
get_task_path Ran
ran; our contract draft was wrong, not the code
openbmb/ultraeval/configs/show_datasets.py
code served (permissive licence) · get_code("a0b5007332000cd3")
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 os
import tempfile
import shutil

# Create a temporary directory structure that matches what get_task_path expects
temp_base = tempfile.mkdtemp()

# Create dataset1/config directory with a config file
dataset1_dir = os.path.join(temp_base, 'dataset1')
config1_dir = os.path.join(dataset1_dir, 'config')
os.makedirs(config1_dir, exist_ok=True)

# Create a config file
config_file1 = os.path.join(config1_dir, 'task_one.json')
with open(config_file1, 'w') as f:
    f.write('{}')

# Create dataset2/config directory with another config file
dataset2_dir = os.path.join(temp_base, 'dataset2')
config2_dir = os.path.join(dataset2_dir, 'config')
os.makedirs(config2_dir, exist_ok=True)

config_file2 = os.path.join(config2_dir, 'task_two_v2.json')
with open(config_file2, 'w') as f:
    f.write('{}')

# Create dataset3 without config directory (should be skipped)
dataset3_dir = os.path.join(temp_base, 'dataset3')
os.makedirs(dataset3_dir, exist_ok=True)

ARGS = {
    'base_dir': temp_base
}
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
thread_function Not yet run
unverified (harvested, no recorded run)
openbmb/ultraeval/models/general_model.py
code served (permissive licence) · get_code("6d5f1662be27380d")
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, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 0 later) and 0 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_task_path 1 other paper
1 dated earlier
2402.04588 (2024-02)
UltraLink : An Open-Source Knowledge-Enhanced Multilingual Supervised Fine-tuning Dataset
get_task_path 1 other paper
1 dated earlier
2402.04588 (2024-02)
UltraLink : An Open-Source Knowledge-Enhanced Multilingual Supervised Fine-tuning Dataset
thread_function 1 other paper
1 dated earlier
2402.04588 (2024-02)
UltraLink : An Open-Source Knowledge-Enhanced Multilingual Supervised Fine-tuning Dataset

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

Evaluation is pivotal for refining Large Language Models (LLMs), pinpointing their capabilities, and guiding enhancements. The rapid development of LLMs calls for a lightweight and easy-to-use framework for swift evaluation deployment. However, considering various implementation details, developing a comprehensive evaluation platform is never easy. Existing platforms are often complex and poorly modularized, hindering seamless incorporation into research workflows. This paper introduces Ul-traEval, a user-friendly evaluation framework characterized by its lightweight nature, comprehensiveness, modularity, and efficiency. We identify and reimplement three core components of model evaluation (models, data, and metrics). The resulting composability allows for the free combination of different models, tasks, prompts, benchmarks, and metrics within a unified evaluation workflow. Additionally, UltraEval supports diverse models owing to a unified HTTP service and provides sufficient inference acceleration. UltraEval is now available for researchers publicly 1 .

For agents

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

get_harvested_code_for_paper("2404.07584")
get_code_for_paper("2404.07584")
have("2404.07584")

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

curl https://syntology.ai/api/ran/2404.07584.json

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

[![Syntology run record](https://syntology.ai/api/ran/2404.07584.svg)](https://syntology.ai/paper/2404.07584)

Connect an agent — have() is free.