SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.06031 · 2026

AP-OOD: Attention Pooling for Out-of-Distribution Detection

Werner Zellinger, Sepp Hochreiter, Daniel Klotz, Claus Hofmann, Bernhard Lehner, Christian Huber

arXiv · PDF · Open in the Atlas

Code that ran

1 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; 0 failures of our instrument; 3 no recorded run), measured 2026-11-22; 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/2602.06031.json · badge: /api/ran/2602.06031.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
ml-jku/ap-ood community
Papers with Code archive
1 ran · 3 no recorded run
FunctionStatusWhere it lives
dataset_collate_fn Ran
ran on a synthesized input
ml-jku/ap-ood/src/ap_ood_experiments/method/ap_ood.py
code served (permissive licence) · get_code("03a218ef6d7f6eb1")
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 function. 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
from torch.utils.data import Dataset, ConcatDataset, Subset

torch.manual_seed(0)

# Create a simple dataset class with a collate_fn
class SimpleDataset(Dataset):
    def __init__(self, size=5, has_collate=False):
        self.size = size
        self.has_collate = has_collate
        if has_collate:
            self.collate_fn = lambda batch: torch.stack([item[0] for item in batch])
        else:
            self.collate_fn = None
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        return torch.randn(4), idx

# Create a dataset without collate_fn
class DatasetNoCollate(Dataset):
    def __init__(self, size=5):
        self.size = size
        self.collate_fn = None
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        return torch.randn(4), idx

# Test case 1: Dataset with collate_fn
dataset_with_collate = SimpleDataset(size=5, has_collate=True)

# Test case 2: Subset wrapping a dataset with collate_fn
base_dataset = SimpleDataset(size=10, has_collate=True)
subset_dataset = Subset(base_dataset, indices=[0, 1, 2])

# Test case 3: ConcatDataset with mixed datasets
dataset1 = DatasetNoCollate(size=3)
dataset2 = SimpleDataset(size=4, has_collate=True)
concat_dataset = ConcatDataset([dataset1, dataset2])

# Test case 4: ConcatDataset with no collate_fn in any dataset
dataset3 = DatasetNoCollate(size=2)
dataset4 = DatasetNoCollate(size=3)
concat_no_collate = ConcatDataset([dataset3, dataset4])

ARGS = {
    "dataset": dataset_with_collate
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
APOODMethod Not yet run
unverified (harvested, no recorded run)
ml-jku/ap-ood/src/ap_ood_experiments/method/ap_ood.py
code served (permissive licence) · get_code("c90e3821e47e638e")
How it ran

no invocation recorded.

Method Not yet run
unverified (harvested, no recorded run)
ml-jku/ap-ood/src/ap_ood_experiments/method/ap_ood.py
code served (permissive licence) · get_code("5de88dc51f406a27")
How it ran

no invocation recorded.

safe_histogram Not yet run
unverified (harvested, no recorded run)
ml-jku/ap-ood/src/ap_ood_experiments/method/ap_ood.py
code served (permissive licence) · get_code("32be8332e47d3ad7")
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, 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 4 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

Out-of-distribution (OOD) detection, which maps high-dimensional data into a scalar OOD score, is critical for the reliable deployment of machine learning models. A key challenge in recent research is how to effectively leverage and aggregate token embeddings from language models to obtain the OOD score. In this work, we propose AP-OOD, a novel OOD detection method for natural language that goes beyond simple average-based aggregation by exploiting token-level information. AP-OOD is a semi-supervised approach that flexibly interpolates between unsupervised and supervised settings, enabling the use of limited auxiliary outlier data. Empirically, AP-OOD sets a new state of the art in OOD detection for text: in the unsupervised setting, it reduces the FPR95 (false positive rate at 95% true positives) from 27.84% to 4.67% on XSUM summarization, and from 77.08% to 70.37% on WMT15 En-Fr translation. Code is available at https://github.com/ml-jku/ap-ood. 1. We propose AP-OOD, an OOD detection approach for natural language that leverages token-level information to detect OOD sequences.

For agents

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

get_harvested_code_for_paper("2602.06031")
get_code_for_paper("2602.06031")
have("2602.06031")

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

curl https://syntology.ai/api/ran/2602.06031.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.06031.svg)](https://syntology.ai/paper/2602.06031)

Connect an agent — have() is free.