SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.13087 · CVPR · 2023

Robust Generalization against Photon-Limited Corruptions via Worst-Case Sharpness Minimization

Bo Han, Chen Gong, Jun Yu, Tongliang Liu, Li Shen, Zhuo Huang, Xiaobo Xia, Bo Du, Miaoxi Zhu

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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; 2 failures of our instrument; 1 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/2303.13087.json · badge: /api/ran/2303.13087.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
feedbackward/addro community
Papers with Code archive
3 ran · 0 no recorded run
zhuohuangai/sharpdro marked official in Papers with Code
Papers with Code archive
0 ran · 1 no recorded run
FunctionStatusWhere it lives
Loss_SoftAD Ran
ran on a synthesized input
feedbackward/addro/addro/setup/losses.py
code served (permissive licence) · get_code("8c43c6c7c8a7a882")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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 torch
from torch import Tensor
from torch.nn import Module, CrossEntropyLoss, BCELoss, L1Loss, Parameter

def get_named_loss(loss_name, loss_paras, reduction):
    if loss_name == 'CrossEntropy':
        loss_fn = CrossEntropyLoss(reduction=reduction)
    elif loss_name == 'BCELoss':
        loss_fn = BCELoss(reduction=reduction)
    elif loss_name == 'L1Loss':
        loss_fn = L1Loss(reduction=reduction)
    else:
        raise ValueError('Unrecognized loss name.')
    return loss_fn

def rho_torch(x):
    return torch.sqrt(1.0 + x ** 2) - 1.0

class Loss_SoftAD(Module):
    """
    Soft ascent-descent (SoftAD), our most basic modified version of
    the flooding algorithm.
    """

    def __init__(self, theta: float, sigma: float, eta: float, loss_name: str, loss_paras: dict):
        super().__init__()
        self.theta = theta
        self.sigma = sigma
        self.eta = eta
        self.loss_fn = get_named_loss(loss_name=loss_name, loss_paras=loss_paras, reduction='none')
        return None

    def forward(self, input: Tensor, target: Tensor):
        theta = self.theta
        sigma = self.sigma + 1e-12
        eta = self.eta
        loss = self.loss_fn(input=input, target=target)
        dispersion = sigma ** 2 * rho_torch((loss - theta) / sigma).mean()
        return eta * theta + dispersion


INIT_ARGS = {
    'theta': 0.5,
    'sigma': 0.2,
    'eta': 1.0,
    'loss_name': 'L1Loss',
    'loss_paras': {}
}

ARGS = {
    'input': torch.randn(2, 4),
    'target': torch.randn(2, 4)
}

if __name__ == '__main__':
    loss_module = Loss_SoftAD(**INIT_ARGS)
    output = loss_module(**ARGS)
    print(f"Output: {output}")
    print(f"Output shape: {output.shape}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
get_named_loss Ran
ran; our contract draft was wrong, not the code
feedbackward/addro/addro/setup/losses.py
code served (permissive licence) · get_code("ef3414913760ce94")
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 CrossEntropyLoss. 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.nn import CrossEntropyLoss, BCELoss, L1Loss

torch.manual_seed(0)

ARGS = {
    'loss_name': 'CrossEntropy',
    'loss_paras': {},
    'reduction': 'mean'
}
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 CrossEntropyLoss — a CrossEntropyLoss output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a CrossEntropyLoss 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
rho_torch Ran
ran; our contract draft was wrong, not the code
feedbackward/addro/addro/setup/losses.py
code served (permissive licence) · get_code("5f9e3f7cfa214e9f")
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 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

torch.manual_seed(0)

# The function rho_torch takes a single parameter x and applies a mathematical operation to it
# x should be a tensor that supports arithmetic operations (sqrt, power, subtraction)

x = torch.randn(2, 4)

ARGS = {
    "x": x
}
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 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
SAM Not yet run
unverified (harvested, no recorded run)
zhuohuangai/sharpdro/sam.py
code served (permissive licence) · get_code("91898e953f3b8aa0")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 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
rho_torch
brick of this paper's assembly: helper
2 other papers
2 later · 1 only as an assembly brick
2310.10006 (2023-10)
Soft ascent-descent as a stable and flexible alternative to flooding
2402.09802 (2024-02)
Criterion Collapse and Loss Distribution Control
Loss_SoftAD
brick of this paper's assembly: entry
1 other paper
1 later
2310.10006 (2023-10)
Soft ascent-descent as a stable and flexible alternative to flooding

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

Robust generalization aims to tackle the most challenging data distributions which are rare in the training set and contain severe noises, i.e., photon-limited corruptions. Common solutions such as distributionally robust optimization (DRO) focus on the worst-case empirical risk to ensure low training error on the uncommon noisy distributions. However, due to the over-parameterized model being optimized on scarce worst-case data, DRO fails to produce a smooth loss landscape, thus struggling on generalizing well to the test set. Therefore, instead of focusing on the worst-case risk minimization, we propose SharpDRO by penalizing the sharpness of the worst-case distribution, which measures the loss changes around the neighbor of learning parameters. Through worst-case sharpness minimization, the proposed method successfully produces a flat loss curve on the corrupted distributions, thus achieving robust generalization. Moreover, by considering whether the distribution annotation is available, we apply SharpDRO to two problem settings and design a worst-case selection process for robust generalization. Theoretically, we show that SharpDRO has a great convergence guarantee. Experimentally, we simulate photon-limited corruptions using CIFAR10/100 and Ima-geNet30 datasets and show that SharpDRO exhibits a strong generalization ability against severe corruptions and exceeds well-known baseline methods with large performance gains.

For agents

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

get_harvested_code_for_paper("2303.13087")
get_code_for_paper("2303.13087")
have("2303.13087")

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

curl https://syntology.ai/api/ran/2303.13087.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.13087.svg)](https://syntology.ai/paper/2303.13087)

Connect an agent — have() is free.