SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2310.09505 · 2023

Advancing Test-Time Adaptation in Wild Acoustic Test Settings

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 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 1 has no contract verdict; 5 failures of our instrument; 2 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/2310.09505.json · badge: /api/ran/2310.09505.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
Waffle-Liu/CEA canonical (classifier label)
classifier basis: census
6 ran · 2 no recorded run
FunctionStatusWhere it lives
copy_model_and_optimizer Ran
ran; our contract draft was wrong, not the code
Waffle-Liu/CEA/utils.py
code served (permissive licence) · get_code("aa08b2e342dc684c")
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 torch
import torch.nn as nn
from torch.optim import SGD

torch.manual_seed(0)

# Build a minimal model
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

model = SimpleModel()

# Build a list of optimizers
optimizer = [
    SGD(model.parameters(), lr=0.01),
    SGD(model.parameters(), lr=0.001)
]

ARGS = {
    "model": model,
    "optimizer": optimizer
}
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 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
create_dataset Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Waffle-Liu/CEA/data.py
code served (permissive licence) · get_code("6e52321ce4391a33")
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 torch
import sys
from unittest.mock import MagicMock, patch

torch.manual_seed(0)

# Create a minimal stand-in Dataset class that satisfies the function's needs
class MockDataset:
    def __init__(self, batch_size, path, name=None, noise_type=None, noise_level=None):
        self.batch_size = batch_size
        self.path = path
        self.name = name
        self.noise_type = noise_type
        self.noise_level = noise_level
        self.data = [(torch.randn(2, 4), torch.randn(2, 3)) for _ in range(5)]
    
    def __len__(self):
        return len(self.data)
    
    def __getitem__(self, idx):
        return self.data[idx]

# Mock the corpus modules to avoid import errors
mock_librispeech = MagicMock()
mock_librispeech.LibriDataset = MockDataset
mock_librispeech.NoisyLibriDataset = MockDataset

mock_chime = MagicMock()
mock_chime.CHiMEDataset = MockDataset

mock_l2arctic = MagicMock()
mock_l2arctic.L2ArcticDataset = MockDataset

mock_dsing = MagicMock()
mock_dsing.DSingDataset = MockDataset

sys.modules['corpus'] = MagicMock()
sys.modules['corpus.librispeech'] = mock_librispeech
sys.modules['corpus.CHiME'] = mock_chime
sys.modules['corpus.l2arctic'] = mock_l2arctic
sys.modules['corpus.DSing'] = mock_dsing

ARGS = {
    'name': 'librispeech',
    'path': '/tmp/librispeech',
    'batch_size': 2,
    'noise_type': None,
    'noise_level': None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
corpus, 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
load_model_and_optimizer Ran
ran; our contract draft was wrong, not the code
Waffle-Liu/CEA/utils.py
code served (permissive licence) · get_code("cead882920870dcb")
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 torch
import torch.nn as nn

torch.manual_seed(0)

# Create a simple model
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

model = SimpleModel()

# Create optimizer list - the function iterates over optimizer with len() and indexing
optimizer = [
    torch.optim.SGD(model.parameters(), lr=0.01),
    torch.optim.Adam(model.parameters(), lr=0.001)
]

# Get initial state dicts
model_state = model.state_dict()
optimizer_state = [opt.state_dict() for opt in optimizer]

# Modify the model and optimizer states slightly to simulate they came from a checkpoint
model.linear.weight.data.fill_(0.5)
for opt in optimizer:
    opt.step()

# Now create fresh instances to load states into
model_fresh = SimpleModel()
optimizer_fresh = [
    torch.optim.SGD(model_fresh.parameters(), lr=0.01),
    torch.optim.Adam(model_fresh.parameters(), lr=0.001)
]

ARGS = {
    "model": model_fresh,
    "optimizer": optimizer_fresh,
    "model_state": model_state,
    "optimizer_state": optimizer_state
}
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 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
preprocess_text Ran
ran; our contract draft was wrong, not the code
Waffle-Liu/CEA/corpus/l2arctic.py
code served (permissive licence) · get_code("e6c3900dd3f5b5a3")
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.

import re
from builtins import str as unicode

def preprocess_text(text):
    text = unicode(text)
    text = text.replace('i.e.', 'that is')
    text = text.replace('e.g.', 'for example')
    text = text.replace('Mr.', 'Mister')
    text = text.replace('Mrs.', 'Mistress')
    text = text.replace('Dr.', 'Doctor')
    text = text.replace('-', ' ')
    text = text.upper()
    text = re.sub("[^ A-Z']", '', text)
    text = ' '.join(text.split())
    return text

ARGS = {
    'text': "Hello, this is a test. Mr. Smith and Dr. Jones met i.e. they encountered each other, e.g., at the conference."
}
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
setup_optimizer Ran
ran; our contract draft was wrong, not the code
Waffle-Liu/CEA/utils.py
code served (permissive licence) · get_code("f5679e2c4c24ad92")
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 torch
import torch.nn as nn

torch.manual_seed(0)

# Create a small model with parameters to optimize
model1 = nn.Linear(4, 3)
model2 = nn.Linear(3, 2)

# Extract parameters from each model
params1 = list(model1.parameters())
params2 = list(model2.parameters())

# Create learning rates corresponding to each parameter group
lrs = [0.001, 0.0005]

ARGS = {
    'params': [params1, params2],
    'lrs': lrs,
    'weight_decay': 0.01,
    'opt_name': 'AdamW'
}
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 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
snr_mixer Ran
ran on a synthesized input
Waffle-Liu/CEA/corpus/audiolib.py
code served (permissive licence) · get_code("482275f8ce08c946")
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 numpy as np

# Create small numpy arrays for clean and noise signals
clean = np.random.randn(100).astype(np.float32)
noise = np.random.randn(100).astype(np.float32)

# SNR value in dB
snr = 10.0

ARGS = {
    "clean": clean,
    "noise": noise,
    "snr": snr
}
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 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
read_text Not yet run
unverified (harvested, no recorded run)
Waffle-Liu/CEA/corpus/CHiME.py
code served (permissive licence) · get_code("8f95ea6583bad9ea")
How it ran

no invocation recorded.

read_text Not yet run
unverified (harvested, no recorded run)
Waffle-Liu/CEA/corpus/librispeech.py
code served (permissive licence) · get_code("0fc0a419d5c391d6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 6 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 3 later) and 2 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
snr_mixer 2 other papers
1 dated earlier, 1 later
2306.01981 (2023-06)
SGEM: Test-Time Adaptation for Automatic Speech Recognition via Sequential-Level Generalized Entropy Minimization
2024.emnlp-main.405 (2024 (year only))
Advancing Test-Time Adaptation in Wild Acoustic Test Settings
copy_model_and_optimizer 1 other paper
1 later
2601.16240 (2026-01)
Test-Time Adaptation for Speech Emotion Recognition
load_model_and_optimizer 1 other paper
1 later
2601.16240 (2026-01)
Test-Time Adaptation for Speech Emotion Recognition
preprocess_text 1 other paper
1 dated earlier
2306.01981 (2023-06)
SGEM: Test-Time Adaptation for Automatic Speech Recognition via Sequential-Level Generalized Entropy Minimization
read_text 1 other paper
1 dated earlier
2306.01981 (2023-06)
SGEM: Test-Time Adaptation for Automatic Speech Recognition via Sequential-Level Generalized Entropy Minimization
read_text 1 other paper
1 dated earlier
2306.01981 (2023-06)
SGEM: Test-Time Adaptation for Automatic Speech Recognition via Sequential-Level Generalized Entropy Minimization

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

Acoustic foundation models, fine-tuned for Automatic Speech Recognition (ASR), suffer from performance degradation in wild acoustic test settings when deployed in real-world scenarios. Stabilizing online Test-Time Adaptation (TTA) under these conditions remains an open and unexplored question. Existing wild vision TTA methods often fail to handle speech data effectively due to the unique characteristics of high-entropy speech frames, which are unreliably filtered out even when containing crucial semantic content. Furthermore, unlike static vision data, speech signals follow short-term consistency, requiring specialized adaptation strategies. In this work, we propose a novel wild acoustic TTA method tailored for ASR fine-tuned acoustic foundation models. Our method, Confidence-Enhanced Adaptation, performs frame-level adaptation using a confidence-aware weight scheme to avoid filtering out essential information in high-entropy frames. Additionally, we apply consistency regularization during test-time optimization to leverage the inherent short-term consistency of speech signals. Our experiments on both synthetic and real-world datasets demonstrate that our approach outperforms existing baselines under various wild acoustic test settings, including Gaussian noise, environmental sounds, accent variations, and sung speech.

For agents

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

get_harvested_code_for_paper("2310.09505")
get_code_for_paper("2310.09505")
have("2310.09505")

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

curl https://syntology.ai/api/ran/2310.09505.json

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

[![Syntology run record](https://syntology.ai/api/ran/2310.09505.svg)](https://syntology.ai/paper/2310.09505)

Connect an agent — have() is free.