SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.06562 · ICLR · 2023

ContraNorm: A Contrastive Learning Perspective on Oversmoothing and Beyond

Yisen Wang, Yifei Wang, Tianqi Du, Xiaojun Guo

arXiv · PDF · Open in the Atlas

Code that ran

1 of 1 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; 0 failures of our instrument; 0 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/2303.06562.json · badge: /api/ran/2303.06562.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
pku-ml/contranorm community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
ContraNorm Ran
ran on a synthesized input
pku-ml/contranorm/ViT_imagenet/models.py
code served (permissive licence) · get_code("09c0ad730c80fc82")
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 3, dtype family float, shape (2, 3, 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 torch.nn as nn

INIT_ARGS = {
    'dim': 4,
    'scale': 0.5,
    'dual_norm': False,
    'pre_norm': False,
    'temp': 1.0,
    'learnable': False,
    'positive': False,
    'identity': False,
}

ARGS = {
    'x': torch.randn(2, 3, 4),
}

if __name__ == '__main__':
    model = nn.Module()
    
    # Import the class
    import sys
    sys.path.insert(0, '.')
    
    # Define ContraNorm inline for testing
    class ContraNorm(nn.Module):

        def __init__(self, dim, scale=0.0, dual_norm=False, pre_norm=False, temp=1.0, learnable=False, positive=False, identity=False):
            super().__init__()
            if learnable and scale > 0:
                import math
                if positive:
                    scale_init = math.log(scale)
                else:
                    scale_init = scale
                self.scale_param = nn.Parameter(torch.empty(dim).fill_(scale_init))
            self.dual_norm = dual_norm
            self.scale = scale
            self.pre_norm = pre_norm
            self.temp = temp
            self.learnable = learnable
            self.positive = positive
            self.identity = identity
            self.layernorm = nn.LayerNorm(dim, eps=1e-06)

        def forward(self, x):
            if self.scale > 0.0:
                xn = nn.functional.normalize(x, dim=2)
                if self.pre_norm:
                    x = xn
                sim = torch.bmm(xn, xn.transpose(1, 2)) / self.temp
                if self.dual_norm:
                    sim = nn.functional.softmax(sim, dim=2) + nn.functional.softmax(sim, dim=1)
                else:
                    sim = nn.functional.softmax(sim, dim=2)
                x_neg = torch.bmm(sim, x)
                if not self.learnable:
                    if self.identity:
                        x = (1 + self.scale) * x - self.scale * x_neg
                    else:
                        x = x - self.scale * x_neg
                else:
                    scale = torch.exp(self.scale_param) if self.positive else self.scale_param
                    scale = scale.view(1, 1, -1)
                    if self.identity:
                        x = scale * x - scale * x_neg
                    else:
                        x = x - scale * x_neg
            x = self.layernorm(x)
            return x

    model = ContraNorm(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Output: {output}")
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 3, dtype family float, shape (2, 3, 4) — shape (2, 3, 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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)

Also present in other papers' code

Of 1 samples lifted from this paper's linked repositories, 1 carry a function identity and 0 do not. Of the 1 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 1 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.

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

Oversmoothing is a common phenomenon in a wide range of Graph Neural Networks (GNNs) and Transformers, where performance worsens as the number of layers increases. Instead of characterizing oversmoothing from the view of complete collapse in which representations converge to a single point, we dive into a more general perspective of dimensional collapse in which representations lie in a narrow cone. Accordingly, inspired by the effectiveness of contrastive learning in preventing dimensional collapse, we propose a novel normalization layer called ContraNorm. Intuitively, ContraNorm implicitly shatters representations in the embedding space, leading to a more uniform distribution and a slighter dimensional collapse. On the theoretical analysis, we prove that ContraNorm can alleviate both complete collapse and dimensional collapse under certain conditions. Our proposed normalization layer can be easily integrated into GNNs and Transformers with negligible parameter overhead. Experiments on various real-world datasets demonstrate the effectiveness of our proposed ContraNorm. Our implementation is available at https://github.com/PKU-ML/ContraNorm.

For agents

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

get_harvested_code_for_paper("2303.06562")
get_code_for_paper("2303.06562")
have("2303.06562")

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

curl https://syntology.ai/api/ran/2303.06562.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.06562.svg)](https://syntology.ai/paper/2303.06562)

Connect an agent — have() is free.