SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2607.02437 · 2026

Extreme Adaptive Transformer for Time Series Forecasting EXTREME ADAPTIVE TRANSFORMER FOR TIME SERIES FORECASTING

Yifan Zhang, Hui Liu, Sanjeev Shrestha

arXiv · PDF · Open in the Atlas

Code that ran

9 of 12 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, 3 violated it and 0 have no contract verdict; 5 failures of our instrument; 3 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/2607.02437.json · badge: /api/ran/2607.02437.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
sanzexstha/Exformer canonical (classifier label)
classifier basis: census
9 ran · 3 no recorded run
FunctionStatusWhere it lives
CORR Ran
ran, violated the contract we drafted
sanzexstha/Exformer/utils/metrics.py
code served (permissive licence) · get_code("e6663d05846f08ea")
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 float32, ndim 0, dtype family float, shape () 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 numpy as np
import torch

torch.manual_seed(0)

# CORR expects two numpy arrays: pred and true
# Looking at the function body:
# - It computes means along axis 0
# - It performs element-wise operations
# - It sums along axis 0
# - It computes a final mean along axis -1
# So we need 2D arrays where axis 0 is batch dimension and axis 1 is feature dimension

# Create small distinct-sized arrays
# Using shape (2, 4) for batch size 2, 4 features
np.random.seed(0)
pred = np.random.randn(2, 4).astype(np.float32)
true = np.random.randn(2, 4).astype(np.float32)

ARGS = {
    "pred": pred,
    "true": true,
}
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 float32, ndim 0, dtype family float, shape () — shape () 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
CORR Ran
ran, violated the contract we drafted
sanzexstha/Exformer/utils/metrics_dan.py
code served (permissive licence) · get_code("03d63d172032b256")
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 float32, ndim 0, dtype family float, shape () 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 numpy as np

# Create small numpy arrays for pred and true
# CORR expects 2D arrays where it computes correlation across the first dimension
# Using distinct sizes: batch=2, features=4
np.random.seed(0)
pred = np.random.randn(2, 4).astype(np.float32)
true = np.random.randn(2, 4).astype(np.float32)

ARGS = {
    "pred": pred,
    "true": true
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float32, ndim 0, dtype family float, shape () — shape () 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
Corr Ran
ran, violated the contract we drafted
sanzexstha/Exformer/utils/metrics.py
code served (permissive licence) · get_code("edd5108dc9b37e58")
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 float64, ndim 0, 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 numpy as np

# Create small numpy arrays for pred and true
# The function operates on numpy arrays with shape (batch_size, num_features)
np.random.seed(0)

pred = np.random.randn(2, 4)  # batch_size=2, num_features=4
true = np.random.randn(2, 4)  # batch_size=2, num_features=4

ARGS = {
    'pred': pred,
    'true': true
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
FFT_for_Period Ran
ran; our contract draft was wrong, not the code
sanzexstha/Exformer/models/my_method/build_model_util.py
code served (permissive licence) · get_code("15503d40fcb7d6f1")
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 tuple, ndim 2, 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 torch
import numpy as np

torch.manual_seed(0)

# x is used with:
# - torch.fft.rfft(x, dim=0) - needs to be a tensor
# - x.shape[0] - needs shape attribute
# - abs(xf).mean(-1)[top_list] - xf comes from rfft, so x should have multiple dimensions
# The function computes FFT along dim=0, then uses shape[0] for period calculation
# x needs to have at least 2 dimensions (the last dimension is used in mean(-1))

x = torch.randn(16, 4)  # shape (16, 4): 16 time steps, 4 features
k = 2

ARGS = {
    "x": x,
    "k": k
}
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 tuple, ndim 2, dtype family float — a tuple 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
all_reduce_mean Ran
ran, honoured the contract we drafted
sanzexstha/Exformer/utils/misc.py
code served (permissive licence) · get_code("3dc19396537db789")
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 float, ndim 0, dtype family float, shape () as a later run recorded it. 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, 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 torch.distributed as dist

torch.manual_seed(0)

# The function all_reduce_mean takes a single parameter 'x'
# Looking at the function body:
# - x is converted to a torch.tensor
# - If world_size > 1, it's moved to cuda and used with dist.all_reduce
# - Otherwise, x is returned as-is
# Since we're CPU-only and dist is not initialized, the function will return x unchanged
# We need to provide a numeric value that can be converted to a tensor

ARGS = {
    "x": 5.0
}
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 float, ndim 0, dtype family float, shape () — shape () 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
labels_to_segments Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sanzexstha/Exformer/models/my_method/dozerformer_EncDec.py
code served (permissive licence) · get_code("5310cdce54653355")
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 Tensor, ndim 3, dtype family float, shape (2, 2, 3) 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
from einops import rearrange, reduce, repeat

torch.manual_seed(0)

# Create a small tensor with shape (B, L, C)
# B=2 (batch size), L=8 (sequence length), C=3 (channels)
# This allows patch_size=4 to divide evenly: 8 // 4 = 2 segments
y = torch.randint(0, 2, (2, 8, 3), dtype=torch.float32)

# patch_size: the length of each patch segment
patch_size = 4

# patch_thres: threshold for counting ones to determine majority label
# With patch_size=4, threshold of 2 means >= 2 ones out of 4 elements
patch_thres = 2

# in_channel is not used in the function body, but it's a parameter
in_channel = None

ARGS = {
    "y": y,
    "patch_size": patch_size,
    "patch_thres": patch_thres,
    "in_channel": in_channel,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, 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 3, dtype family float, shape (2, 2, 3) — shape (2, 2, 3) 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
load_config Ran
ran; our contract draft was wrong, not the code
sanzexstha/Exformer/utils/ext_utils.py
code served (permissive licence) · get_code("735aba507a02ddfe")
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 tempfile
import os

# Create a temporary config file
config_content = """# This is a comment
key1=value1
key2=value2
enabled=true
learning_rate=0.001
"""

with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt') as f:
    f.write(config_content)
    temp_filepath = f.name

try:
    ARGS = {
        'filepath': temp_filepath
    }
finally:
    pass
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
normalize_diff_with_stats Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sanzexstha/Exformer/utils/ext_utils.py
code served (permissive licence) · get_code("276f14f028e3ea4f")
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 ndarray, ndim 1, dtype family float, shape (6,) 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 numpy as np
import torch

# Set seed for reproducibility
np.random.seed(0)

# Create sample sensor data - a list of numeric values
sensor_data_val = [0.5, 1.2, 0.8, 1.5, 0.3, 2.1]

# Create mean and std for normalization
# These should be numpy arrays or scalars that work with subtraction and division
mean = np.array(0.5)
std = np.array(0.3)

ARGS = {
    "sensor_data_val": sensor_data_val,
    "mean": mean,
    "std": std
}
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 ndarray, ndim 1, dtype family float, shape (6,) — shape (6,) 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
r_log_std_normalization Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
sanzexstha/Exformer/utils/ext_utils.py
code served (permissive licence) · get_code("7a0a7a0da04b61d4")
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 tuple. 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 numpy as np
import torch
import torch.utils.data as data

torch.manual_seed(0)

# The function expects sensor_data_val to be:
# - indexable with [1:] slicing
# - iterable with len()
# - elements comparable with > 0
# - arithmetic operations supported

# Create a small numpy array as sensor data
sensor_data = np.array([0.5, 1.2, -0.3, 2.1, 0.8])

ARGS = {
    "sensor_data_val": sensor_data
}
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 tuple — a tuple 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
MAE Not yet run
unverified (harvested, no recorded run)
sanzexstha/Exformer/utils/metrics_dan.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from shangzongjiang/Ada-MSHyper, is pointer only (licence: NONE) · get_code("75f44993b096bf76")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from shangzongjiang/Ada-MSHyper, and GitHub detected no licence in that repository when it was checked. 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. No invocation recorded.

RSE Not yet run
unverified (harvested, no recorded run)
sanzexstha/Exformer/utils/metrics.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from shangzongjiang/Ada-MSHyper, is pointer only (licence: NONE) · get_code("b40a11875ebd0cd2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from shangzongjiang/Ada-MSHyper, and GitHub detected no licence in that repository when it was checked. 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. No invocation recorded.

get_grad_norm_ Not yet run
unverified (harvested, no recorded run)
sanzexstha/Exformer/utils/misc.py
code served (permissive licence) · get_code("a93537b210192d0c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 8 are also present in other papers whose code we hold (8 with at least one paper dated earlier, 1 the same month, 0 later) and 4 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
all_reduce_mean 25 other papers
24 dated earlier, 1 the same month
2105.12723 (2021-05)
Nested Hierarchical Transformer: Towards Accurate, Data-Efficient and Interpretable Visual Understanding
2204.09435 (2022-04)
Hephaestus: A large scale multitask dataset towards InSAR understanding
2211.12941 (2022-11)
EurNet: Efficient Multi-Range Relational Modeling of Spatial Multi-Relational Data
and 22 more not named here
get_grad_norm_ 21 other papers
21 dated earlier
2308.00566 (2023-08)
Stochastic positional embeddings improve masked image modeling
2310.12979 (2023-10)
Predicting a Protein's Stability under a Million Mutations
2311.13627 (2023-11)
Vamos: Versatile Action Models for Video Understanding
and 18 more not named here
RSE 18 other papers
18 dated earlier
2202.01381 (2022-02)
ETSformer: Exponential Smoothing Transformers for Time-series Forecasting
2203.05167 (2022-03)
TiSAT: Time Series Anomaly Transformer
2207.10941 (2022-07)
Respecting Time Series Properties Makes Deep Time Series Forecasting Perfect
and 15 more not named here
MAE 17 other papers
17 dated earlier
2202.01381 (2022-02)
ETSformer: Exponential Smoothing Transformers for Time-series Forecasting
2203.05167 (2022-03)
TiSAT: Time Series Anomaly Transformer
2207.10941 (2022-07)
Respecting Time Series Properties Makes Deep Time Series Forecasting Perfect
and 14 more not named here
CORR 13 other papers
13 dated earlier
2202.01381 (2022-02)
ETSformer: Exponential Smoothing Transformers for Time-series Forecasting
2203.05167 (2022-03)
TiSAT: Time Series Anomaly Transformer
2207.10941 (2022-07)
Respecting Time Series Properties Makes Deep Time Series Forecasting Perfect
and 10 more not named here
Corr 2 other papers
2 dated earlier
2308.11946 (2023-08)
Multi-scale Transformer Pyramid Networks for Multivariate Time Series Forecasting
2405.11333 (2024-05)
GinAR: An End-To-End Multivariate Time Series Forecasting Model Suitable for Variable Missing
CORR 1 other paper
1 dated earlier
2308.11946 (2023-08)
Multi-scale Transformer Pyramid Networks for Multivariate Time Series Forecasting
FFT_for_Period 1 other paper
1 dated earlier
2308.11946 (2023-08)
Multi-scale Transformer Pyramid Networks for Multivariate Time Series Forecasting

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

Time series forecasting remains challenging when the underlying data contain rare but critical extreme events. This issue is particularly important in hydrologic forecasting, where streamflow distributions are often highly skewed and extreme peaks can have substantial impacts on flood monitoring, water resource management, and early warning systems. Although Transformer-based forecasting models have achieved strong performance by modeling long-range temporal dependencies, they typically treat all time points uniformly and may therefore underrepresent rare extreme patterns. In this paper, we propose the Extreme-Adaptive Transformer (Exformer), a forecasting framework designed to explicitly model temporal dependencies involving both normal and extreme events. Exformer introduces an extreme-adaptive attention mechanism composed of three sparse components: Local, Stride, and Extreme. The Local and Stride components capture short-term and periodic temporal dependencies, respectively, while the Extreme component selectively models event-aware dependencies between normal and extreme streamflow patterns. Experiments on four real-world hydrologic streamflow datasets show that Exformer achieves superior 3-day forecasting performance compared with state-of-the-art baselines. Our findings demonstrate that explicitly incorporating extreme-aware attention improves the forecasting capacity of Transformer models on imbalanced time series with rare but consequential events. Code is available at https://github.com/sanzexstha/Exformer.

For agents

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

get_harvested_code_for_paper("2607.02437")
get_code_for_paper("2607.02437")
have("2607.02437")

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

curl https://syntology.ai/api/ran/2607.02437.json

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

[![Syntology run record](https://syntology.ai/api/ran/2607.02437.svg)](https://syntology.ai/paper/2607.02437)

Connect an agent — have() is free.