SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.03184 · ICML · 2020

Time-aware Large Kernel Convolutions

Vasileios Lioutas, Yuhong Guo

arXiv · PDF · Open in the Atlas

Code that ran

6 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 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; 6 failures of our instrument; 1 no recorded run), measured 2026-11-21; 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/2002.03184.json · badge: /api/ran/2002.03184.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
lioutasb/TaLKConvolutions canonical (classifier label)
classifier basis: census
4 ran · 1 no recorded run
lioutasb/talkconvolutions canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
FunctionStatusWhere it lives
Embedding Ran
ran; our contract draft was wrong, not the code · constructed an object; no result was computed
lioutasb/TaLKConvolutions/talkconv/talkconv_fairseq/talkconv.py
pointer only (licence: MIT); the stored copy get_code serves, from instance-wise-ordered-transformer/iot, is pointer only (licence: NOASSERTION) · get_code("96cbb5e9ca5b6be0")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from instance-wise-ordered-transformer/iot, and GitHub could not classify the licence of that repository (NOASSERTION). 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.

Linear Ran
ran; our contract draft was wrong, not the code · constructed an object; no result was computed
lioutasb/TaLKConvolutions/talkconv/talkconv_fairseq/talkconv.py
pointer only (licence: MIT); the stored copy get_code serves, from instance-wise-ordered-transformer/iot, is pointer only (licence: NOASSERTION) · get_code("8cd8cb0d1e9e63e4")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from instance-wise-ordered-transformer/iot, and GitHub could not classify the licence of that repository (NOASSERTION). 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.

average_checkpoints Ran
ran; our contract draft was wrong, not the code
lioutasb/talkconvolutions/utils/average_checkpoints.py
code served (permissive licence) · get_code("93ba0aaaba0bfada")
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) 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 collections
import tempfile
import os

torch.manual_seed(0)

# Create temporary checkpoint files
temp_dir = tempfile.mkdtemp()

# Create first checkpoint
checkpoint1 = {
    'model': collections.OrderedDict([
        ('layer1.weight', torch.randn(4, 3)),
        ('layer1.bias', torch.randn(4)),
        ('layer2.weight', torch.randn(5, 4)),
    ])
}
checkpoint1_path = os.path.join(temp_dir, 'checkpoint1.pt')
torch.save(checkpoint1, checkpoint1_path)

# Create second checkpoint with same structure
checkpoint2 = {
    'model': collections.OrderedDict([
        ('layer1.weight', torch.randn(4, 3)),
        ('layer1.bias', torch.randn(4)),
        ('layer2.weight', torch.randn(5, 4)),
    ])
}
checkpoint2_path = os.path.join(temp_dir, 'checkpoint2.pt')
torch.save(checkpoint2, checkpoint2_path)

# Create third checkpoint with same structure
checkpoint3 = {
    'model': collections.OrderedDict([
        ('layer1.weight', torch.randn(4, 3)),
        ('layer1.bias', torch.randn(4)),
        ('layer2.weight', torch.randn(5, 4)),
    ])
}
checkpoint3_path = os.path.join(temp_dir, 'checkpoint3.pt')
torch.save(checkpoint3, checkpoint3_path)

ARGS = {
    'inputs': [checkpoint1_path, checkpoint2_path, checkpoint3_path]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
gelu Ran
ran; our contract draft was wrong, not the code
lioutasb/TaLKConvolutions/talkconv/talkconv_fairseq/talkconv.py
code served (permissive licence) · get_code("bddb1c75a5038e42")
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 2, dtype family float, shape (2, 4) 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 math

torch.manual_seed(0)

# The gelu function takes a single parameter 'x' which is used in mathematical operations
# It should be a tensor that supports arithmetic operations and torch.tanh
x = torch.randn(2, 4)

ARGS = {
    "x": x
}
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 Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
last_n_checkpoints Ran
ran; our contract draft was wrong, not the code
lioutasb/talkconvolutions/utils/average_checkpoints.py
code served (permissive licence) · get_code("8dd1def34e4485c2")
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) 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
import torch

# Create a temporary directory with checkpoint files
temp_dir = tempfile.mkdtemp()

# Create checkpoint files matching the expected patterns
# For update_based=True: checkpoint_<step>_<update>.pt
# For update_based=False: checkpoint<step>.pt

# Create some checkpoint files for testing
checkpoint_files = [
    "checkpoint1.pt",
    "checkpoint2.pt",
    "checkpoint3.pt",
    "checkpoint5.pt",
]

for filename in checkpoint_files:
    filepath = os.path.join(temp_dir, filename)
    torch.save({"dummy": "data"}, filepath)

# Also create some update-based checkpoints
update_based_files = [
    "checkpoint_100_1.pt",
    "checkpoint_100_2.pt",
    "checkpoint_100_3.pt",
]

for filename in update_based_files:
    filepath = os.path.join(temp_dir, filename)
    torch.save({"dummy": "data"}, filepath)

ARGS = {
    "paths": [temp_dir],
    "n": 2,
    "update_based": False,
    "upper_bound": None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
last_n_checkpoints Ran
ran; our contract draft was wrong, not the code
lioutasb/TaLKConvolutions/utils/average_checkpoints.py
pointer only (licence: MIT); the stored copy get_code serves, from ubc-nlp/llmas, is pointer only (licence: NOASSERTION) · get_code("99de76104bad1fc9")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from ubc-nlp/llmas, and GitHub could not classify the licence of that repository (NOASSERTION). 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.

average_checkpoints Not yet run
unverified (harvested, no recorded run)
lioutasb/TaLKConvolutions/utils/average_checkpoints.py
code served (permissive licence) · get_code("7cc2e301e30389d1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 5 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 5 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
Linear 58 other papers
6 dated earlier, 52 later · 5 only as an assembly brick
1409.0473 (2014-09)
Neural Machine Translation by Jointly Learning to Align and Translate
1706.03762 (2017-06)
Attention Is All You Need
1904.01681 (2019-04)
Augmented Neural ODEs
and 55 more not named here
Embedding 29 other papers
2 dated earlier, 27 later
1904.01681 (2019-04)
Augmented Neural ODEs
1906.12183 (2019-06)
Neural ODEs as the Deep Limit of ResNets with constant weights
2003.09229 (2020-03)
Learning to Encode Position for Transformer with Continuous Dynamical Model
and 26 more not named here
average_checkpoints 4 other papers
4 later
2008.07772 (2020-08)
Very Deep Transformers for Neural Machine Translation
2210.07535 (2022-10)
AutoMoE: Heterogeneous Mixture-of-Experts with Adaptive Computation for Efficient Neural Machine Translation
2310.16712 (2023-10)
LLM Performance Predictors are good initializers for Architecture Search
and 1 more not named here
last_n_checkpoints 3 other papers
3 later
2008.07772 (2020-08)
Very Deep Transformers for Neural Machine Translation
2210.07535 (2022-10)
AutoMoE: Heterogeneous Mixture-of-Experts with Adaptive Computation for Efficient Neural Machine Translation
2310.16712 (2023-10)
LLM Performance Predictors are good initializers for Architecture Search
gelu 1 other paper
1 later
2207.08625 (2022-07)
Unifying Event Detection and Captioning as Sequence Generation via Pre-Training

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

To date, most state-of-the-art sequence modeling architectures use attention to build generative models for language based tasks. Some of these models use all the available sequence tokens to generate an attention distribution which results in time complexity of O(n 2 ). Alternatively, they utilize depthwise convolutions with softmax normalized kernels of size k acting as a limited-window self-attention, resulting in time complexity of O(k•n). In this paper, we introduce Time-aware Large Kernel (TaLK) Convolutions, a novel adaptive convolution operation that learns to predict the size of a summation kernel instead of using a fixed-sized kernel matrix. This method yields a time complexity of O(n), effectively making the sequence encoding process linear to the number of tokens. We evaluate the proposed method on large-scale standard machine translation, abstractive summarization and language modeling datasets and show that TaLK Convolutions constitute an efficient improvement over other attention/convolution based approaches.

For agents

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

get_harvested_code_for_paper("2002.03184")
get_code_for_paper("2002.03184")
have("2002.03184")

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

curl https://syntology.ai/api/ran/2002.03184.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.03184.svg)](https://syntology.ai/paper/2002.03184)

Connect an agent — have() is free.