SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.06774 · AAAI · 2020

Residual Continual Learning

Janghyeon Lee, Junmo Kim, Donggyu Joo, Hyeong Hong

arXiv · PDF · Open in the Atlas

Code that ran

4 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 1 failure of our instrument; 4 no recorded run), measured 2026-11-21; 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/2002.06774.json · badge: /api/ran/2002.06774.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
parvex/residual-continual-learning-benchmark community
Papers with Code archive
4 ran · 4 no recorded run
FunctionStatusWhere it lives
AverageMeter Ran
ran on a synthesized input · constructed an object; no result was computed
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("b208280d7046fd72")
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 an instance of its own class, AverageMeter: the run constructed the object and did not call it, so no result was computed. 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

class AverageMeter(object):
    """Computes and stores the average and current value"""

    def __init__(self):
        self.reset()

    def reset(self):
        self.val = 0
        self.avg = 0
        self.sum = 0
        self.count = 0

    def update(self, val, n=1):
        self.val = val
        self.sum += val * n
        self.count += n
        self.avg = float(self.sum) / self.count


# Test the AverageMeter with sample setup
meter = AverageMeter()
meter.update(5.0, n=1)
meter.update(10.0, n=2)

ARGS = {}
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 AverageMeter — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a AverageMeter 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
PreActResNet_cifar Ran
ran on a synthesized input · constructed an object; no result was computed
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("1699ef27e2fb05cf")
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 an instance of its own class, PreActResNet_cifar: the run constructed the object and did not call it, so no result was computed. 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)

# Define a minimal ResNet block that matches the interface expected
class PreActBlock(nn.Module):
    expansion = 1
    
    def __init__(self, in_planes, out_planes, stride=1, droprate=0):
        super(PreActBlock, self).__init__()
        self.bn1 = nn.BatchNorm2d(in_planes)
        self.conv1 = nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride, padding=1, bias=False)
        self.bn2 = nn.BatchNorm2d(out_planes)
        self.conv2 = nn.Conv2d(out_planes, out_planes, kernel_size=3, stride=1, padding=1, bias=False)
        
        self.shortcut = nn.Sequential()
        if stride != 1 or in_planes != out_planes:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False)
            )
    
    def forward(self, x):
        out = F.relu(self.bn1(x))
        out = self.conv1(out)
        out = F.relu(self.bn2(out))
        out = self.conv2(out)
        out += self.shortcut(x)
        return out

# Import F for use in the block
import torch.nn.functional as F

ARGS = {
    "block": PreActBlock,
    "num_blocks": [2, 2, 2],
    "filters": [16, 32, 64],
    "num_classes": 10,
    "droprate": 0.0
}
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 PreActResNet_cifar — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a PreActResNet_cifar 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
Timer Ran
ran on a synthesized input · constructed an object; no result was computed
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("a734a8b4d53236df")
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 an instance of its own class, Timer: the run constructed the object and did not call it, so no result was computed. 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 time

class Timer(object):
    """
    """

    def __init__(self):
        self.reset()

    def reset(self):
        self.interval = 0
        self.time = time.time()

    def value(self):
        return time.time() - self.time

    def tic(self):
        self.time = time.time()

    def toc(self):
        self.interval = time.time() - self.time
        self.time = time.time()
        return self.interval


ARGS = {}
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 Timer — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Timer 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
accumulate_acc Ran
ran; our contract draft was wrong, not the code
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("e88bef5f49e74025")
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 AccuracyMeter. 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)

# Small stand-in class for the meter that tracks accuracy
class AccuracyMeter:
    def __init__(self):
        self.correct = 0
        self.total = 0
    
    def update(self, acc, n):
        self.correct += acc * n / 100.0
        self.total += n
    
    def reset(self):
        self.correct = 0
        self.total = 0

# Create test data
batch_size = 2
num_classes = 4

# output: dict with either 'All' key or task-specific keys
# Case 1: Using 'All' key - simpler case
output = {
    'All': torch.randn(batch_size, num_classes)
}

# target: tensor of class indices
target = torch.tensor([0, 2])

# task: list indicating which task each sample belongs to
task = [0, 0]

# meter: AccuracyMeter instance
meter = AccuracyMeter()

ARGS = {
    'output': output,
    'target': target,
    'task': task,
    'meter': meter
}
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 AccuracyMeter — a AccuracyMeter output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a AccuracyMeter 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
CombinedResNet Not yet run
unverified (harvested, no recorded run)
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("ac4fe8de4b29c7a2")
How it ran

no invocation recorded.

NormalNN Not yet run
unverified (harvested, no recorded run)
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("15b764b71fe6c2a2")
How it ran

no invocation recorded.

ResCL Not yet run
unverified (harvested, no recorded run)
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("5be0c9e3f02178ff")
How it ran

no invocation recorded.

accuracy Not yet run
unverified (harvested, no recorded run)
parvex/residual-continual-learning-benchmark/agents/regularization.py
code served (permissive licence) · get_code("0c4bbcb160a04955")
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, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 6 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
Timer
brick of this paper's assembly: helper
2 other papers
2 later · 2 only as an assembly brick
2106.09701 (2021-06)
Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning
2410.20444 (2024-10)
Vector Quantization Prompting for Continual Learning
accuracy
brick of this paper's assembly: helper
2 other papers
2 later · 1 only as an assembly brick
2106.09701 (2021-06)
Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning
2410.20444 (2024-10)
Vector Quantization Prompting for Continual Learning

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

We propose a novel continual learning method called Residual Continual Learning (ResCL). Our method can prevent the catastrophic forgetting phenomenon in sequential learning of multiple tasks, without any source task information except the original network. ResCL reparameterizes network parameters by linearly combining each layer of the original network and a fine-tuned network; therefore, the size of the network does not increase at all. To apply the proposed method to general convolutional neural networks, the effects of batch normalization layers are also considered. By utilizing residuallearning-like reparameterization and a special weight decay loss, the trade-off between source and target performance is effectively controlled. The proposed method exhibits state-ofthe-art performance in various continual learning scenarios.

For agents

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

get_harvested_code_for_paper("2002.06774")
get_code_for_paper("2002.06774")
have("2002.06774")

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

curl https://syntology.ai/api/ran/2002.06774.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.06774.svg)](https://syntology.ai/paper/2002.06774)

Connect an agent — have() is free.