SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2306.09347 · NeurIPS · 2023

Segment Any Point Cloud Sequences by Distilling Vision Foundation Models

Ziwei Liu, Kai Chen, Liang Pan, Lingdong Kong, Jun Cen, Runnan Chen, Youquan Liu, Wenwei Zhang

arXiv · PDF · Open in the Atlas

Code that ran

18 of 24 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 1 violated it and 9 have no contract verdict; 7 failures of our instrument; 6 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/2306.09347.json · badge: /api/ran/2306.09347.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
IDEA-Research/OpenSeeD canonical (classifier label)
classifier basis: census
10 ran · 4 no recorded run
microsoft/X-Decoder canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
subake/DAPS3D community
Papers with Code archive
3 ran · 2 no recorded run
valeoai/SLidR community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
Bottleneck Ran
ran on a synthesized input · constructed an object; no result was computed
subake/DAPS3D/train/tasks/semantic/modules/ddrnet_23_slim_salsa_da.py
code served (permissive licence) · get_code("ab585f4a45f6186a")
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, Bottleneck: 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)

# Define a minimal downsample module that matches how it's used in forward()
class MinimalDownsample(nn.Module):
    def forward(self, x):
        # The downsample is called with x and should return a tensor of matching shape
        # For Bottleneck with expansion=2, we need output channels = inplanes * 2
        # If stride > 1, spatial dimensions are reduced
        return x

# Create instances for the arguments
inplanes = 4
planes = 3
stride = 1
no_relu = True

# Create a downsample module
# The downsample needs to map from inplanes to planes * expansion channels
# and handle stride if needed
downsample = nn.Sequential(
    nn.Conv2d(inplanes, planes * 2, kernel_size=1, stride=stride, bias=False),
    nn.BatchNorm2d(planes * 2, momentum=0.1)
)

ARGS = {
    'inplanes': inplanes,
    'planes': planes,
    'stride': stride,
    'downsample': downsample,
    'no_relu': no_relu
}
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 Bottleneck — 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 Bottleneck 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
DAPPM Ran
ran on a synthesized input · constructed an object; no result was computed
subake/DAPS3D/train/tasks/semantic/modules/ddrnet_23_slim_salsa_da.py
code served (permissive licence) · get_code("98542be0286d7f06")
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, DAPPM: 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)

# Define ARGS dict with parameters for DAPPM.__init__
ARGS = {
    'inplanes': 64,
    'branch_planes': 32,
    'outplanes': 128,
}

# Create the module instance to verify it works
model = nn.Module.__new__(type('DAPPM', (nn.Module,), {}))

# Test with a small input tensor
# DAPPM expects input of shape (batch, channels, height, width)
x = torch.randn(2, ARGS['inplanes'], 16, 20)

# Verify the dimensions make sense for the forward pass
print(f"Input shape: {x.shape}")
print(f"inplanes: {ARGS['inplanes']}")
print(f"branch_planes: {ARGS['branch_planes']}")
print(f"outplanes: {ARGS['outplanes']}")
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 DAPPM — 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 DAPPM 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
VisionTransformer Ran
ran on a synthesized input
valeoai/SLidR/model/modules/dino/vision_transformer.py
pointer only (licence: NOASSERTION) · get_code("684c765a98595395")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, valeoai/SLidR (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

batch_dice_loss Ran
ran; our contract draft was wrong, not the code
IDEA-Research/OpenSeeD/openseed/modules/matcher.py
code served (permissive licence) · get_code("bc2cb481a75c370d")
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 2, dtype family float, shape (2, 2) 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 small tensors for inputs and targets
# inputs: arbitrary shape float tensor (predictions)
# targets: same shape as inputs, binary classification labels (0 or 1)
# Using shape (2, 4) - batch size 2, 4 features
inputs = torch.randn(2, 4, dtype=torch.float32)
targets = torch.randint(0, 2, (2, 4), dtype=torch.float32)

ARGS = {
    'inputs': inputs,
    'targets': targets,
}
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 2, dtype family float, shape (2, 2) — shape (2, 2) 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
bbox_postprocess Ran
ran on a synthesized input
IDEA-Research/OpenSeeD/openseed/modules/postprocessing.py
code served (permissive licence) · get_code("d3d81dca00185824")
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 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)

# Create a small result tensor: [xc, yc, w, h] in range [0, 1]
# Shape: (batch_size=2, 4)
result = torch.rand(2, 4)

# input_size: [height, width] of the input to the model
input_size = (480, 640)

# img_size: [height, width] of the actual image
img_size = (360, 480)

# output_height and output_width: dimensions for final scaling
output_height = 720
output_width = 960

ARGS = {
    "result": result,
    "input_size": input_size,
    "img_size": img_size,
    "output_height": output_height,
    "output_width": output_width,
}
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 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
build_decoder Ran
ran on a synthesized input
microsoft/X-Decoder/modeling/body/decoder/build.py
code served (permissive licence) · get_code("e3b73f26b6efd55f")
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 MinimalDecoder. 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 unittest.mock import MagicMock

torch.manual_seed(0)

# Create a minimal stand-in for model registry functions
def is_model(name):
    return name in ["decoder_v1", "decoder_v2"]

def model_entrypoints(name):
    class MinimalDecoder(nn.Module):
        def __init__(self, config, *args, **kwargs):
            super().__init__()
            self.config = config
            self.linear = nn.Linear(4, 3)
        
        def forward(self, x):
            return self.linear(x)
    
    return MinimalDecoder

# Patch the functions used by build_decoder
import sys
current_module = sys.modules[__name__]
current_module.is_model = is_model
current_module.model_entrypoints = model_entrypoints

# Create a config dict with the required structure
config = {
    'MODEL': {
        'DECODER': {
            'NAME': 'decoder_v1'
        }
    },
    'HIDDEN_DIM': 64,
    'NUM_LAYERS': 2
}

ARGS = {
    'config': config
}
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 MinimalDecoder — a MinimalDecoder output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MinimalDecoder 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
dice_loss Ran
ran; our contract draft was wrong, not the code
IDEA-Research/OpenSeeD/openseed/modules/criterion.py
code served (permissive licence) · get_code("89f75e54ff128be0")
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 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

torch.manual_seed(0)

# Create small tensors for inputs and targets
# inputs should be arbitrary shape that can be flattened
inputs = torch.randn(2, 4, requires_grad=False)  # batch_size=2, features=4
targets = torch.randint(0, 2, (2, 4), dtype=torch.float32)  # same shape as inputs, binary labels

# num_masks is a float representing the number of masks for normalization
num_masks = 2.0

ARGS = {
    "inputs": inputs,
    "targets": targets,
    "num_masks": num_masks,
}
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 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: none by the harness; any seed is in the setup block
filter_images_with_only_crowd_annotations Ran
ran; our contract draft was wrong, not the code
microsoft/X-Decoder/datasets/build.py
code served (permissive licence) · get_code("a21a61e41d986b6a")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 logging

# Set up logging to avoid warnings
logging.basicConfig(level=logging.INFO)

# Create sample dataset_dicts in Detectron2 format
dataset_dicts = [
    {
        'image_id': 1,
        'file_name': 'image1.jpg',
        'annotations': [
            {'id': 1, 'category_id': 1, 'bbox': [10, 20, 100, 150], 'iscrowd': 0},
            {'id': 2, 'category_id': 2, 'bbox': [50, 60, 80, 120], 'iscrowd': 0},
        ]
    },
    {
        'image_id': 2,
        'file_name': 'image2.jpg',
        'annotations': [
            {'id': 3, 'category_id': 1, 'bbox': [30, 40, 120, 160], 'iscrowd': 1},
        ]
    },
    {
        'image_id': 3,
        'file_name': 'image3.jpg',
        'annotations': [
            {'id': 4, 'category_id': 2, 'bbox': [15, 25, 110, 155], 'iscrowd': 0},
        ]
    },
    {
        'image_id': 4,
        'file_name': 'image4.jpg',
        'annotations': []
    },
]

# Create sample dataset_names
dataset_names = ['coco_train', 'coco_val']

ARGS = {
    'dataset_dicts': dataset_dicts,
    'dataset_names': dataset_names,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
generate_regular_grid_point_coords Ran
ran, honoured the contract we drafted
IDEA-Research/OpenSeeD/openseed/modules/point_features.py
code served (permissive licence) · get_code("42ab1750d0229462")
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 3, dtype family float, shape (3, 16, 2) 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.nn import functional as F

torch.manual_seed(0)

ARGS = {
    "R": 3,
    "side_size": 4,
    "device": torch.device("cpu"),
}
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 3, dtype family float, shape (3, 16, 2) — shape (3, 16, 2) 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
is_model Ran
ran, violated the contract we drafted
IDEA-Research/OpenSeeD/openseed/architectures/registry.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from ux-decoder/dinov, is pointer only (licence: NONE) · get_code("86c6e09f9f6f6637")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from ux-decoder/dinov, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

point_sample Ran
ran; our contract draft was wrong, not the code
IDEA-Research/OpenSeeD/openseed/modules/point_features.py
code served (permissive licence) · get_code("39e4af93b70bcddf")
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 3, dtype family float, shape (2, 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

torch.manual_seed(0)

# input: (N, C, H, W) - batch_size=2, channels=4, height=8, width=6
input_tensor = torch.randn(2, 4, 8, 6)

# point_coords: (N, P, 2) - batch_size=2, num_points=3, coordinates=2
# Values in [0, 1] range as required by the function
point_coords = torch.rand(2, 3, 2)

ARGS = {
    "input": input_tensor,
    "point_coords": point_coords,
}
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 3, dtype family float, shape (2, 4, 3) — shape (2, 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: none by the harness; any seed is in the setup block
register_decoder Ran
ran on a synthesized input
microsoft/X-Decoder/modeling/body/decoder/registry.py
code served (permissive licence) · get_code("e3c3001fcab78f5c")
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 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
import torch.nn as nn

def dummy_decoder():
    """A simple dummy function to use as the decoder function"""
    pass

# Set the module name to simulate a proper module structure
dummy_decoder.__module__ = 'mymodels.decoder_module'

ARGS = {
    'fn': dummy_decoder
}
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 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
register_model Ran
ran; our contract draft was wrong, not the code
IDEA-Research/OpenSeeD/openseed/architectures/registry.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from ux-decoder/dinov, is pointer only (licence: NONE) · get_code("be2fbb8bf98a15f7")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from ux-decoder/dinov, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

register_norm_module Ran
ran on a synthesized input
microsoft/X-Decoder/utils/model.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from UARK-AICV/OpenFusion, is pointer only (licence: NONE) · get_code("e4f6733fa1ab1999")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from UARK-AICV/OpenFusion, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

segmenthead Ran
ran on a synthesized input · constructed an object; no result was computed
subake/DAPS3D/train/tasks/semantic/modules/ddrnet_23_slim_salsa_da.py
code served (permissive licence) · get_code("093af22c7f46b460")
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, segmenthead: 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)

# The segmenthead module takes:
# - inplanes: number of input channels
# - interplanes: number of intermediate channels
# - outplanes: number of output channels
# - scale_factor: optional scaling factor for interpolation

ARGS = {
    'inplanes': 16,
    'interplanes': 32,
    'outplanes': 8,
    'scale_factor': 2
}
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 segmenthead — 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 segmenthead 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
sem_seg_postprocess Ran
ran on a synthesized input
IDEA-Research/OpenSeeD/openseed/modules/postprocessing.py
code served (permissive licence) · get_code("0e8074c3542f9ab5")
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 3, dtype family float, shape (4, 12, 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

torch.manual_seed(0)

# Create a semantic segmentation prediction tensor (C, H, W)
# C=4 classes, H=8 prediction height, W=6 prediction width
result = torch.randn(4, 8, 6)

# Image size that the segmentor takes as input (height, width)
img_size = (7, 5)

# Desired output resolution
output_height = 12
output_width = 10

ARGS = {
    "result": result,
    "img_size": img_size,
    "output_height": output_height,
    "output_width": output_width,
}
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 3, dtype family float, shape (4, 12, 10) — shape (4, 12, 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: none by the harness; any seed is in the setup block
sigmoid_ce_loss Ran
ran; our contract draft was wrong, not the code
IDEA-Research/OpenSeeD/openseed/modules/criterion.py
code served (permissive licence) · get_code("d0c61e8dba511aa3")
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 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
import torch.nn.functional as F

torch.manual_seed(0)

# Create small tensors for inputs and targets
# inputs: predictions (logits), arbitrary shape
# targets: binary labels (0 or 1), same shape as inputs
batch_size = 2
num_elements = 4

inputs = torch.randn(batch_size, num_elements, dtype=torch.float32)
targets = torch.randint(0, 2, (batch_size, num_elements), dtype=torch.float32)

# num_masks: used as divisor in the loss computation
num_masks = 2.0

ARGS = {
    "inputs": inputs,
    "targets": targets,
    "num_masks": num_masks,
}
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 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: none by the harness; any seed is in the setup block
sigmoid_focal_loss Ran
ran; our contract draft was wrong, not the code
IDEA-Research/OpenSeeD/openseed/modules/criterion.py
code served (permissive licence) · get_code("5c0711aada67957e")
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 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
import torch.nn.functional as F

torch.manual_seed(0)

# Create small tensors for inputs and targets
# inputs: predictions (logits) - shape (2, 4) for batch_size=2, num_predictions=4
inputs = torch.randn(2, 4, requires_grad=True)

# targets: binary labels - shape (2, 4) matching inputs
targets = torch.randint(0, 2, (2, 4), dtype=torch.float32)

# num_boxes: number of boxes for normalization
num_boxes = 2

# alpha and gamma are already float parameters with defaults
alpha = 0.25
gamma = 2.0

ARGS = {
    "inputs": inputs,
    "targets": targets,
    "num_boxes": num_boxes,
    "alpha": alpha,
    "gamma": gamma,
}
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 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: none by the harness; any seed is in the setup block
DualResNet Not yet run
unverified (harvested, no recorded run)
subake/DAPS3D/train/tasks/semantic/modules/ddrnet_23_slim_salsa_da.py
code served (permissive licence) · get_code("b635e36051ab5ddb")
How it ran

no invocation recorded.

TransformerStage Not yet run
unverified (harvested, no recorded run)
subake/DAPS3D/train/tasks/semantic/modules/ddrnet_23_slim_salsa_da.py
code served (permissive licence) · get_code("7c9b14c925230f96")
How it ran

no invocation recorded.

batch_sigmoid_ce_loss Not yet run
unverified (harvested, no recorded run)
IDEA-Research/OpenSeeD/openseed/modules/matcher.py
code served (permissive licence) · get_code("1edd24985036b0bf")
How it ran

no invocation recorded.

build_model Not yet run
unverified (harvested, no recorded run)
IDEA-Research/OpenSeeD/openseed/architectures/build.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from ux-decoder/dinov, is pointer only (licence: NONE) · get_code("c67e4671ab9094aa")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from ux-decoder/dinov, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

model_entrypoints Not yet run
unverified (harvested, no recorded run)
IDEA-Research/OpenSeeD/openseed/architectures/registry.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from ux-decoder/dinov, is pointer only (licence: NONE) · get_code("0c3902b80e9968b5")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from ux-decoder/dinov, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

multi_head_attention_forward Not yet run
unverified (harvested, no recorded run)
IDEA-Research/OpenSeeD/openseed/modules/attention.py
code served (permissive licence) · get_code("22a94fc69aed9aca")
How it ran

no invocation recorded.

Also present in other papers' code

Of 24 samples lifted from this paper's linked repositories, 24 carry a function identity and 0 do not. Of the 24 distinct functions with an identity, 14 are also present in other papers whose code we hold (10 with at least one paper dated earlier, 5 the same month, 14 later) and 10 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
sigmoid_focal_loss 56 other papers
28 dated earlier, 1 the same month, 24 later, 3 same year, order not determinable · 1 only as an assembly brick
2005.12872 (2020-05)
End-to-End Object Detection with Transformers
2010.04159 (2020-10)
Deformable DETR: Deformable Transformers for End-to-End Object Detection
2103.11731 (2021-03)
Meta-DETR: Image-Level Few-Shot Object Detection with Inter-Class Correlation Exploitation
and 53 more not named here
dice_loss 46 other papers
12 dated earlier, 4 the same month, 27 later, 3 same year, order not determinable
2112.01527 (2021-12)
Masked-attention Mask Transformer for Universal Image Segmentation
2112.10764 (2021-12)
Mask2Former for Video Instance Segmentation
2203.03605 (2022-03)
DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection
and 43 more not named here
sigmoid_ce_loss 45 other papers
11 dated earlier, 4 the same month, 27 later, 3 same year, order not determinable
2112.01527 (2021-12)
Masked-attention Mask Transformer for Universal Image Segmentation
2112.10764 (2021-12)
Mask2Former for Video Instance Segmentation
2203.03605 (2022-03)
DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection
and 42 more not named here
batch_dice_loss 44 other papers
10 dated earlier, 4 the same month, 27 later, 3 same year, order not determinable
2112.01527 (2021-12)
Masked-attention Mask Transformer for Universal Image Segmentation
2112.10764 (2021-12)
Mask2Former for Video Instance Segmentation
2203.03605 (2022-03)
DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection
and 41 more not named here
batch_sigmoid_ce_loss 44 other papers
10 dated earlier, 4 the same month, 27 later, 3 same year, order not determinable
2112.01527 (2021-12)
Masked-attention Mask Transformer for Universal Image Segmentation
2112.10764 (2021-12)
Mask2Former for Video Instance Segmentation
2203.03605 (2022-03)
DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection
and 41 more not named here
is_model 22 other papers
12 dated earlier, 9 later, 1 same year, order not determinable
1409.4842 (2014-09)
Going deeper with convolutions
1512.00567 (2015-12)
Rethinking the Inception Architecture for Computer Vision
1608.06993 (2016-08)
Densely Connected Convolutional Networks
and 19 more not named here
point_sample 6 other papers
1 dated earlier, 5 later
2103.06564 (2021-03)
PointFlow: Flowing Semantics Through Points for Aerial Image Segmentation
2311.14671 (2023-11)
SEGIC: Unleashing the Emergent Correspondence for In-Context Segmentation
2312.09158 (2023-12)
General Object Foundation Model for Images and Videos at Scale
and 3 more not named here
filter_images_with_only_crowd_annotations 5 other papers
5 later
2311.00571 (2023-11)
LLaVA-Interactive: An All-in-One Demo for Image Chat, Segmentation, Generation and Editing
2311.01373 (2023-11)
Recognize Any Regions
01071 (2024 (year only))
VISAGE: Video Instance Segmentation with Appearance-Guided Enhancement
and 2 more not named here
register_norm_module 4 other papers
1 dated earlier, 3 later
2305.10355 (2023-05)
Evaluating Object Hallucination in Large Vision-Language Models
2310.03923 (2023-10)
Open-Fusion: Real-time Open-Vocabulary 3D Mapping and Queryable Scene Representation
2311.00571 (2023-11)
LLaVA-Interactive: An All-in-One Demo for Image Chat, Segmentation, Generation and Editing
and 1 more not named here
model_entrypoints 2 other papers
1 dated earlier, 1 later
2103.15808 (2021-03)
CvT: Introducing Convolutions to Vision Transformers
2311.13601 (2023-11)
Visual In-Context Prompting
register_model 2 other papers
1 dated earlier, 1 later
2103.15808 (2021-03)
CvT: Introducing Convolutions to Vision Transformers
2311.13601 (2023-11)
Visual In-Context Prompting
build_model 1 other paper
1 later
2311.13601 (2023-11)
Visual In-Context Prompting
generate_regular_grid_point_coords 1 other paper
1 later
2312.09158 (2023-12)
General Object Foundation Model for Images and Videos at Scale
multi_head_attention_forward 1 other paper
1 later
2312.09158 (2023-12)
General Object Foundation Model for Images and Videos at Scale

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

Recent advancements in vision foundation models (VFMs) have opened up new possibilities for versatile and efficient visual perception. In this work, we introduce Seal, a novel framework that harnesses VFMs for segmenting diverse automotive point cloud sequences. Seal exhibits three appealing properties: i) Scalability: VFMs are directly distilled into point clouds, obviating the need for annotations in either 2D or 3D during pretraining. ii) Consistency: Spatial and temporal relationships are enforced at both the camera-to-LiDAR and point-to-segment regularization stages, facilitating cross-modal representation learning. iii) Generalizability: Seal enables knowledge transfer in an off-the-shelf manner to downstream tasks involving diverse point clouds, including those from real/synthetic, low/highresolution, large/small-scale, and clean/corrupted datasets. Extensive experiments conducted on eleven different point cloud datasets showcase the effectiveness and superiority of Seal. Notably, Seal achieves a remarkable 45.0% mIoU on nuScenes after linear probing, surpassing random initialization by 36.9% mIoU and outperforming prior arts by 6.1% mIoU. Moreover, Seal demonstrates significant performance gains over existing methods across 20 different few-shot fine-tuning tasks on all eleven tested point cloud datasets. The code is available at this link 2 .

For agents

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

get_harvested_code_for_paper("2306.09347")
get_code_for_paper("2306.09347")
have("2306.09347")

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

curl https://syntology.ai/api/ran/2306.09347.json

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

[![Syntology run record](https://syntology.ai/api/ran/2306.09347.svg)](https://syntology.ai/paper/2306.09347)

Connect an agent — have() is free.