SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.03390 · 2026

From Vicious to Virtuous Cycles: Synergistic Representation Learning for Unsupervised Video Object-Centric Learning

Jae-Pil Heo, Wonjun Moon, Hyun Seong

arXiv · PDF · Open in the Atlas

Code that ran

1 of 3 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; 2 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/2602.03390.json · badge: /api/ran/2602.03390.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
hynnsk/SRL community
Papers with Code archive
1 ran · 2 no recorded run
FunctionStatusWhere it lives
SupConLoss Ran
ran on a synthesized input
hynnsk/SRL/srl/losses.py
code served (permissive licence) · get_code("81aee15db4dfebcb")
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 nn

class SupConLoss(nn.Module):
    """Supervised Contrastive Learning: https://arxiv.org/pdf/2004.11362.pdf.
    It also supports the unsupervised contrastive loss in SimCLR"""

    def __init__(self, temperature=0.07, base_temperature=0.07):
        super(SupConLoss, self).__init__()
        self.temperature = temperature
        self.base_temperature = base_temperature

    def forward(self, features):
        """Compute loss for model. If both `labels` and `mask` are None,
        it degenerates to SimCLR unsupervised loss:
        https://arxiv.org/pdf/2002.05709.pdf

        Args:
            features: hidden vector of shape [bsz, n_views, ...].
            labels: ground truth of shape [bsz].
            mask: contrastive mask of shape [bsz, bsz], mask_{i,j}=1 if sample j
                has the same class as sample i. Can be asymmetric.
        Returns:
            A loss scalar.
        """
        device = features.device
        features = features.squeeze(0)
        T, B, D = features.size()
        mask = torch.kron(torch.eye(B, dtype=torch.bool, device=device), torch.ones((T, T), dtype=torch.bool, device=device))
        mask_size = mask.size(0)
        ignore = mask.clone()
        for i in range(B):
            start = i * T
            idx = torch.arange(start, start + T - 1, device=device)
            next_idx = idx + 1
            ignore[idx, next_idx] = False
            ignore[next_idx, idx] = False
        ignore = ~ignore
        mask = mask.float()
        ignore = ignore.float()
        mask = mask * ignore
        features_flat = features.permute(1, 0, 2).reshape(B * T, D)
        anchor_dot_contrast = torch.div(torch.matmul(features_flat, features_flat.T), self.temperature)
        logits_max, _ = torch.max(anchor_dot_contrast, dim=1, keepdim=True)
        logits = anchor_dot_contrast - logits_max.detach()
        logits_mask = torch.scatter(torch.ones_like(mask), 1, torch.arange(mask_size, device=device).view(-1, 1), 0)
        mask = mask * logits_mask
        ignore_mask = ignore * logits_mask
        exp_logits = torch.exp(logits) * ignore_mask
        log_prob = logits - torch.log(exp_logits.sum(1, keepdim=True))
        mask_pos_pairs = mask.sum(1)
        mask_pos_pairs = torch.where(mask_pos_pairs < 1e-06, 1, mask_pos_pairs)
        mean_log_prob_pos = (mask * log_prob).sum(1) / mask_pos_pairs
        loss = -(self.temperature / self.base_temperature) * mean_log_prob_pos
        loss = loss.view(mask_size).mean()
        return loss


INIT_ARGS = {
    'temperature': 0.07,
    'base_temperature': 0.07,
}

ARGS = {
    'features': torch.randn(1, 2, 3, 4),
}

if __name__ == '__main__':
    loss_fn = SupConLoss(**INIT_ARGS)
    output = loss_fn(**ARGS)
    print(f"Loss output: {output}")
    print(f"Output shape: {output.shape}")
    print(f"Output type: {type(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 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)
Loss Not yet run
unverified (harvested, no recorded run)
hynnsk/SRL/srl/losses.py
code served (permissive licence) · get_code("69b03cb1053aa22e")
How it ran

no invocation recorded.

Slot_Slot_Contrastive_Loss Not yet run
unverified (harvested, no recorded run)
hynnsk/SRL/srl/losses.py
code served (permissive licence) · get_code("5d3aa80f1dc902bb")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Unsupervised object-centric learning models, particularly slot-based architectures, have shown great promise in decomposing complex scenes. However, their reliance on reconstruction-based training creates a fundamental conflict between the sharp, high-frequency attention maps of the encoder and the spatially consistent but blurry reconstruction maps of the decoder. We identify that this discrepancy gives rise to a vicious cycle: the noisy feature map from the encoder forces the decoder to average over possibilities and produce even blurrier outputs, while the gradient computed from blurry reconstruction maps lacks high-frequency details necessary to supervise encoder features. To break this cycle, we introduce Synergistic Representation Learning (SRL) that establishes a virtuous cycle where the encoder and decoder mutually refine one another. SRL leverages the encoder's sharpness to deblur the semantic boundary within the decoder output, while exploiting the decoder's spatial consistency to denoise the encoder's features. This mutual refinement process is stabilized by a warm-up phase with a slot regularization objective that initially allocates distinct entities per slot. By bridging the representational gap between the encoder and decoder, SRL achieves state-of-the-art results on video object-centric learning benchmarks. Codes are available at github.com/hynnsk/SRL.

For agents

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

get_harvested_code_for_paper("2602.03390")
get_code_for_paper("2602.03390")
have("2602.03390")

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

curl https://syntology.ai/api/ran/2602.03390.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.03390.svg)](https://syntology.ai/paper/2602.03390)

Connect an agent — have() is free.