SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.13566 · NeurIPS · 2020

DISK: Learning local features with policy gradient

Pascal Fua, Michał Tyszkiewicz, Eduard Trulls

arXiv · PDF · Open in the Atlas

Code that ran

8 of 9 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (4 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 6 have no contract verdict; 1 failure 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/2006.13566.json · badge: /api/ran/2006.13566.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
fabio-sim/LightGlue-ONNX community
Papers with Code archive
8 ran · 1 no recorded run
FunctionStatusWhere it lives
Conv Ran
ran on a synthesized input · constructed an object; no result was computed
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("2117c48ae9c7855b")
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, Conv: 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)

ARGS = {
    "in_": 4,
    "out_": 8,
    "size": 3,
    "skip_norm_and_gate": False,
}
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 Conv — the run constructed an instance of its own class and did not call it; no shape is 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
ThinUnetDownBlock Ran
ran on a synthesized input · constructed an object; no result was computed
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("04dca05d9283642e")
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, ThinUnetDownBlock: 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)

ARGS = {
    "in_": 3,
    "out_": 16,
    "size": 5,
    "is_first": False,
}
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 ThinUnetDownBlock — the run constructed an instance of its own class and did not call it; no shape is 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
ThinUnetUpBlock Ran
ran on a synthesized input · constructed an object; no result was computed
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("44727952e7c8b71b")
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, ThinUnetUpBlock: 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)

# Build a minimal ThinUnetUpBlock instance
# Parameters: bottom_: int, horizontal_: int, out_: int, size: int=5

ARGS = {
    "bottom_": 4,
    "horizontal_": 3,
    "out_": 5,
    "size": 5,
}
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 ThinUnetUpBlock — 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 ThinUnetUpBlock 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
TrivialDownsample Ran
ran on a synthesized input
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("27dad2b9963df316")
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 4, dtype family float, shape (2, 4, 4, 3) 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.functional as F
from torch import nn

class TrivialDownsample(nn.Module):

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        return F.avg_pool2d(x, 2)


INIT_ARGS = {}

ARGS = {
    'x': torch.randn(2, 4, 8, 6)
}

if __name__ == '__main__':
    model = TrivialDownsample(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    print(f"Output shape: {output.shape}")
    print(f"Output:\n{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 4, dtype family float, shape (2, 4, 4, 3) — shape (2, 4, 4, 3) 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)
TrivialUpsample Ran
ran on a synthesized input
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("03ca43e2a2d9c9d4")
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 4, dtype family float, shape (2, 4, 6, 10) 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.functional as F
from torch import nn

class TrivialUpsample(nn.Module):

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        return F.interpolate(x, scale_factor=2, mode='bilinear', align_corners=False)


INIT_ARGS = {}

ARGS = {
    'x': torch.randn(2, 4, 3, 5)
}

if __name__ == '__main__':
    model = TrivialUpsample(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    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 4, dtype family float, shape (2, 4, 6, 10) — shape (2, 4, 6, 10) 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)
Unet Ran
ran on a synthesized input · constructed an object; no result was computed
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("7b149b05ea2f9297")
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, Unet: 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)

# Analyze the Unet.__init__ signature:
# - in_features: int (default=1)
# - up: list[int] (default=[])
# - down: list[int] (default=[])
# - size: int (default=5)

# From the code, we see:
# 1. `down` must be 1 item longer than `up` (line: if not len(down) == len(up) + 1)
# 2. down_dims = [in_features, *down] - so down is concatenated with in_features
# 3. path_down iterates through pairwise(down_dims) - each down block connects consecutive dims
# 4. bot_dims = [down[-1], *up] - starts with last element of down
# 5. hor_dims = down_dims[-2::-1] - reversed down_dims excluding the last element
# 6. path_up zips bot_dims, hor_dims, up with strict=False

# Let's create a valid configuration:
# down = [16, 32, 64] (length 3)
# up = [32, 16] (length 2, which is len(down) - 1 = 3 - 1 = 2) ✓

ARGS = {
    'in_features': 1,
    'up': [32, 16],
    'down': [16, 32, 64],
    'size': 5,
}
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 Unet — 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 Unet 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
heatmap_to_keypoints Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("d1954d52abca6433")
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 tuple. 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.functional as F

torch.manual_seed(0)

# Create a small heatmap tensor: batch_size=2, channels=1, height=4, width=3
heatmap = torch.rand(2, 1, 4, 3)

# Number of top keypoints to extract
n = 5

# Window size for max pooling
window_size = 5

ARGS = {
    "heatmap": heatmap,
    "n": n,
    "window_size": window_size,
}
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 tuple — a tuple output has no array shape of its own; none is 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
shape_as_tensor Ran
ran, honoured the contract we drafted
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("efdda0ff601d2d62")
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 Tensor, ndim 1, dtype family int, shape (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)

# Create a small test tensor with distinct dimensions to probe shape later
tensor = torch.randn(2, 4, 3, 5)

ARGS = {
    "tensor": tensor
}
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 Tensor, ndim 1, dtype family int, shape (4,) — shape (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
DISK Not yet run
unverified (harvested, no recorded run)
fabio-sim/LightGlue-ONNX/lightglue_dynamo/models/disk/disk.py
code served (permissive licence) · get_code("d7045b1f4cbcf91c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 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 9 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

Local feature frameworks are difficult to learn in an end-to-end fashion, due to the discreteness inherent to the selection and matching of sparse keypoints. We introduce DISK (DIScrete Keypoints), a novel method that overcomes these obstacles by leveraging principles from Reinforcement Learning (RL), optimizing end-to-end for a high number of correct feature matches. Our simple yet expressive probabilistic model lets us keep the training and inference regimes close, while maintaining good enough convergence properties to reliably train from scratch. Our features can be extracted very densely while remaining discriminative, challenging commonly held assumptions about what constitutes a good keypoint, as showcased in Fig. 1, and deliver state-of-the-art results on three public benchmarks.

For agents

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

get_harvested_code_for_paper("2006.13566")
get_code_for_paper("2006.13566")
have("2006.13566")

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

curl https://syntology.ai/api/ran/2006.13566.json

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

[![Syntology run record](https://syntology.ai/api/ran/2006.13566.svg)](https://syntology.ai/paper/2006.13566)

Connect an agent — have() is free.