SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.00712 · NeurIPS · 2024

ChaosBench: A Multi-Channel, Physics-Based Benchmark for Subseasonal-to-Seasonal Climate Prediction

Aditya Grover, Sungduk Yu, Julius Busecke, Pierre Gentine, Tung Nguyen, Juan Nathaniel, Yongquan Qu

arXiv · PDF · Open in the Atlas

Code that ran

11 of 11 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, 2 violated it and 0 have no contract verdict; 9 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/2402.00712.json · badge: /api/ran/2402.00712.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
leap-stc/ChaosBench canonical (classifier label)
classifier basis: census
9 ran · 0 no recorded run
ecmwf-lab/ai-models canonical (classifier label)
classifier basis: census
2 ran · 0 no recorded run
FunctionStatusWhere it lives
cast_tuple Ran
ran, violated the contract we drafted
leap-stc/ChaosBench/chaosbench/models/vit.py
pointer only (licence: GPL-3.0) · get_code("1f7c550e65142ef8")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

conv2d_bn_relu Ran
ran; our contract draft was wrong, not the code
leap-stc/ChaosBench/chaosbench/models/ae.py
pointer only (licence: GPL-3.0) · get_code("92f1a3fa77bf2251")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

conv_block Ran
ran; our contract draft was wrong, not the code
leap-stc/ChaosBench/chaosbench/models/cnn.py
pointer only (licence: GPL-3.0) · get_code("143f63c99701b6a9")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

convert_time Ran
ran; our contract draft was wrong, not the code
leap-stc/ChaosBench/chaosbench/utils.py
pointer only (licence: GPL-3.0) · get_code("3cd9a84568633114")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

deconv_relu Ran
ran; our contract draft was wrong, not the code
leap-stc/ChaosBench/chaosbench/models/ae.py
pointer only (licence: GPL-3.0) · get_code("451d321a21300bea")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

deconv_sigmoid Ran
ran; our contract draft was wrong, not the code
leap-stc/ChaosBench/chaosbench/models/ae.py
pointer only (licence: GPL-3.0) · get_code("1d3a3f04669fabe7")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

exists Ran
ran, violated the contract we drafted
leap-stc/ChaosBench/chaosbench/models/vit.py
pointer only (licence: GPL-3.0); the stored copy get_code serves, from ThomasMrY/VCT, is pointer only (licence: NONE) · get_code("aa5486a3650902d8")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

interpolate Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
leap-stc/ChaosBench/eval_direct.py
pointer only (licence: GPL-3.0) · get_code("b12e0af56d8cccb1")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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.

peek Ran
ran; our contract draft was wrong, not the code
ecmwf-lab/ai-models/src/ai_models/checkpoint.py
code served (permissive licence) · get_code("9dd689a970ea2365")
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 torch
import os
import pickle
import zipfile
import tempfile
import io

# Create a temporary zip file with a valid data.pkl inside
temp_dir = tempfile.mkdtemp()
zip_path = os.path.join(temp_dir, 'test.zip')

# Create test data to pickle
test_data = {
    'model': {'weights': [1.0, 2.0, 3.0]},
    'config': {'learning_rate': 0.001, 'batch_size': 32},
    'values': [1, 2, 3, 'test', None, True, 3.14],
    'nested': {
        'list': [1, 2, {'inner': 'value'}],
        'tuple': (1, 2, 3)
    }
}

# Write to zip file
with zipfile.ZipFile(zip_path, 'w') as zf:
    pkl_buffer = io.BytesIO()
    pickle.Pickler(pkl_buffer).dump(test_data)
    pkl_buffer.seek(0)
    zf.writestr('data.pkl', pkl_buffer.getvalue())

ARGS = {
    'path': zip_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — 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
tidy Ran
ran; our contract draft was wrong, not the code
ecmwf-lab/ai-models/src/ai_models/checkpoint.py
code served (permissive licence) · get_code("93ab76178df79a49")
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 torch
import numpy as np

torch.manual_seed(0)

# The function 'tidy' recursively processes various data types:
# - dict: recursively tidy all values
# - list: recursively tidy all elements
# - tuple: recursively tidy all elements
# - None: return None
# - int, float, str, bool: return as-is
# - anything else: return as-is

# Create test data with various nested structures
test_dict = {
    "tensor": torch.randn(2, 4),
    "nested_dict": {
        "value": 42,
        "list": [1, 2.5, "string", True, None]
    }
}

test_list = [
    torch.randn(2, 4),
    {"key": "value"},
    (1, 2, 3),
    None,
    42,
    "test"
]

test_tuple = (
    torch.randn(2, 4),
    [1, 2, 3],
    {"nested": True},
    None
)

test_primitive_int = 10
test_primitive_float = 3.14
test_primitive_str = "hello"
test_primitive_bool = True
test_primitive_none = None

# The function takes a single parameter 'x' which can be any of the above types
ARGS = {
    "x": {
        "tensor": torch.randn(2, 4),
        "nested_dict": {
            "value": 42,
            "list": [1, 2.5, "string", True, None]
        },
        "tuple_data": (1, 2, torch.randn(3, 5)),
    }
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — 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
upconv_block Ran
ran; our contract draft was wrong, not the code
leap-stc/ChaosBench/chaosbench/models/cnn.py
pointer only (licence: GPL-3.0) · get_code("362233a242a189a5")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, leap-stc/ChaosBench, is not in the permissive set Syntology serves inline. 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 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 3 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 1 the same month, 3 later) and 8 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
exists 486 other papers
293 dated earlier, 15 the same month, 167 later, 9 same year, order not determinable, 2 undated · 78 only as an assembly brick
1603.08155 (2016-03)
Perceptual Losses for Real-Time Style Transfer and Super-Resolution
1611.07004 (2016-11)
Image-to-Image Translation with Conditional Adversarial Networks
1804.04637 (2018-04)
EMBER: An Open Dataset for Training Static PE Malware Machine Learning Models
and 483 more not named here
cast_tuple 7 other papers
6 dated earlier, 1 later · 1 only as an assembly brick
2102.13419 (2021-02)
Iterative SE(3)-Transformers
2105.12723 (2021-05)
Nested Hierarchical Transformer: Towards Accurate, Data-Efficient and Interpretable Visual Understanding
2105.15203 (2021-05)
SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers
and 4 more not named here
convert_time 1 other paper
1 later
2502.19750 (2025-02)
CirT: Global Subseasonal-to-Seasonal Forecasting with Geometry-inspired Transformer

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

Accurate prediction of climate in the subseasonal-to-seasonal scale is crucial for disaster preparedness and robust decision making amidst climate change. Yet, forecasting beyond the weather timescale is challenging because it deals with problems other than initial condition, including boundary interaction, butterfly effect, and our inherent lack of physical understanding. At present, existing benchmarks tend to have shorter forecasting range of up-to 15 days, do not include a wide range of operational baselines, and lack physics-based constraints for explainability. Thus, we propose ChaosBench, a challenging benchmark to extend the predictability range of data-driven weather emulators to S2S timescale. First, ChaosBench is comprised of variables beyond the typical surface-atmospheric ERA5 to also include ocean, ice, and land reanalysis products that span over 45 years to allow for full Earth system emulation that respects boundary conditions. We also propose physics-based, in addition to deterministic and probabilistic metrics, to ensure a physically-consistent ensemble that accounts for butterfly effect. Furthermore, we evaluate on a diverse set of physics-based forecasts from four national weather agencies as baselines to our data-driven counterpart such as ViT/ClimaX, PanguWeather, GraphCast, and FourCastNetV2. Overall, we find methods originally developed for weather-scale applications fail on S2S task: their performance simply collapse to an unskilled climatology. Nonetheless, we outline and demonstrate several strategies that can extend the predictability range of existing weather emulators, including the use of ensembles, robust control of error propagation, and the use of physics-informed models. Our benchmark, datasets, and instructions are available at https://leap-stc.github.io/ChaosBench.

For agents

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

get_harvested_code_for_paper("2402.00712")
get_code_for_paper("2402.00712")
have("2402.00712")

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

curl https://syntology.ai/api/ran/2402.00712.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.00712.svg)](https://syntology.ai/paper/2402.00712)

Connect an agent — have() is free.