SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2210.01917 · ECCV · 2022

Differentiable Raycasting for Self-supervised Occupancy Forecasting

Deva Ramanan, David Held, Achal Dave, Peiyun Hu, Tarasha Khurana, Jason Ziglar

arXiv · PDF · Open in the Atlas

Code that ran

19 of 20 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; 1 honoured the contract we drafted, 1 violated it and 12 have no contract verdict; 5 failures 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/2210.01917.json · badge: /api/ran/2210.01917.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
tarashakhurana/emergent-occ-forecasting canonical (classifier label)
classifier basis not recorded on the link
19 ran · 1 no recorded run
FunctionStatusWhere it lives
CollateFn Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/data.py
code served (permissive licence) · get_code("18273d36f07fdecf")
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 dict. 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

torch.manual_seed(0)

# Create a batch of examples
batch = [
    {
        'scene_token': 'scene_0',
        'sample_data_token': 'sample_0',
        'input_points': torch.randn(3, 2),
        'sampled_trajectories_fine': torch.randn(4, 5),
        'sampled_trajectories': torch.randn(4, 5),
        'drive_command': torch.randn(2),
        'output_origin': torch.randn(2),
        'output_points': torch.randn(2, 2),
        'gt_trajectory': torch.randn(4, 2),
        'obj_boxes': torch.randn(3, 7),
        'obj_shadows': torch.randn(3, 4),
        'fvf_maps': torch.randn(5, 6),
    },
    {
        'scene_token': 'scene_1',
        'sample_data_token': 'sample_1',
        'input_points': torch.randn(2, 2),
        'sampled_trajectories_fine': torch.randn(4, 5),
        'sampled_trajectories': torch.randn(4, 5),
        'drive_command': torch.randn(2),
        'output_origin': torch.randn(2),
        'output_points': torch.randn(3, 2),
        'gt_trajectory': torch.randn(4, 2),
        'obj_boxes': torch.randn(3, 7),
        'obj_shadows': torch.randn(3, 4),
        'fvf_maps': torch.randn(5, 6),
    },
]

ARGS = {
    'batch': batch,
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
accumulate_scores Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/evaluation.py
code served (permissive licence) · get_code("f7291472e2400f04")
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 ndarray, ndim 1, dtype family float, shape (0,) 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 numpy as np
import numba

# Read the function to understand what arguments it needs:
# - iou: 2D array with shape (num_gt, num_pred)
# - pred_scores: 1D array with shape (num_pred,)
# - gt_flag: 1D array with shape (num_gt,) with values -1, 0, or 1
# - pred_flag: 1D array with shape (num_pred,) with values -1 or 1
# - iou_threshold: scalar float

# Create small test data
num_gt = 3
num_pred = 4

# iou matrix: (num_gt, num_pred) - intersection over union scores
iou = np.array([
    [0.8, 0.3, 0.1, 0.5],
    [0.2, 0.9, 0.4, 0.1],
    [0.1, 0.2, 0.7, 0.3]
], dtype=np.float64)

# pred_scores: confidence scores for each prediction
pred_scores = np.array([0.9, 0.85, 0.75, 0.6], dtype=np.float64)

# gt_flag: ground truth flags (-1 = ignore, 0 = negative, 1 = positive)
gt_flag = np.array([1, 0, 1], dtype=np.int32)

# pred_flag: prediction flags (-1 = ignore, 1 = valid)
pred_flag = np.array([1, 1, 1, -1], dtype=np.int32)

# iou_threshold: minimum IoU to consider a match
iou_threshold = 0.5

ARGS = {
    'iou': iou,
    'pred_scores': pred_scores,
    'gt_flag': gt_flag,
    'pred_flag': pred_flag,
    'iou_threshold': iou_threshold
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (0,) — shape (0,) 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
compute_split_parts Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/eval_utils.py
code served (permissive licence) · get_code("b4e126077df53f1a")
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 int; 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 torch
import numpy as np

torch.manual_seed(0)

ARGS = {
    "num_samples": 10,
    "num_parts": 3
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — 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 int — 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
conv3x3 Ran
ran; our contract draft was wrong, not the code
tarashakhurana/emergent-occ-forecasting/model.py
code served (permissive licence) · get_code("caeaa72cca0b0383")
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 Conv2d. 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_channels": 3,
    "out_channels": 64,
    "bias": 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 Conv2d — a Conv2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Conv2d 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
convex_hull_intersection Ran
ran; our contract draft was wrong, not the code
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/rotate_iou_cpu_eval.py
pointer only (licence: MIT); the stored copy get_code serves, from hailanyi/CPD, is pointer only (licence: NONE) · get_code("31c0e5a46ae6999a")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from hailanyi/CPD, 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.

deconv3x3 Ran
ran; our contract draft was wrong, not the code
tarashakhurana/emergent-occ-forecasting/model.py
code served (permissive licence) · get_code("aa8ad83ca9710526")
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 ConvTranspose2d. 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_channels": 3,
    "out_channels": 64,
    "stride": 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 ConvTranspose2d — a ConvTranspose2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ConvTranspose2d 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
distance_filter Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/eval_utils.py
code served (permissive licence) · get_code("b8b1c25361b6acd3")
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 ndarray, ndim 1, dtype family float, shape (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 numpy as np

# Construct a small boxes array with shape (2, 6)
# The function uses boxes[:, 0:3] which are the first 3 columns (coordinates)
# and expects boxes.shape[0] for the number of boxes
boxes = np.array([
    [10.0, 5.0, 8.0, 1.0, 2.0, 3.0],
    [40.0, 30.0, 20.0, 4.0, 5.0, 6.0]
], dtype=np.float32)

# Test with level 0
level = 0

ARGS = {
    "boxes": boxes,
    "level": level
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (2,) — shape (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
div_up Ran
ran, honoured the contract we drafted
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/iou_utils.py
pointer only (licence: MIT); the stored copy get_code serves, from datongjie/bevspread, is pointer only (licence: NONE) · get_code("20fbfa2279e1ce4b")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from datongjie/bevspread, 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.

draw_object_shadow Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/rasterize.py
code served (permissive licence) · get_code("0f466ea6d440fa49")
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 ndarray, ndim 2, dtype family float, shape (704, 400) 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 numpy as np

class Box:
    """Stand-in class for box objects with bottom_corners method"""
    def __init__(self, corners):
        self.corners = corners
    
    def bottom_corners(self):
        """Return bottom corners as (2, N) array of [Y, X] coordinates"""
        return self.corners

# Create test boxes with bottom corners
# Each box needs 2 rows (Y, X) and some number of corner points
box1_corners = np.array([
    [10.0, 15.0, 20.0, 10.0],  # Y coordinates of 4 corners
    [5.0, 8.0, 12.0, 3.0]       # X coordinates of 4 corners
])
box2_corners = np.array([
    [30.0, 35.0, 40.0, 30.0],  # Y coordinates of 4 corners
    [15.0, 18.0, 22.0, 13.0]   # X coordinates of 4 corners
])

box1 = Box(box1_corners)
box2 = Box(box2_corners)

ARGS = {
    "boxes": [box1, box2],
    "x0": 0.0,
    "y0": 0.0,
    "xlim": [-40.0, 40.0, 0.2],
    "ylim": [-70.4, 70.4, 0.2]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, skimage — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (704, 400) — shape (704, 400) 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
find_nearest Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/data.py
code served (permissive licence) · get_code("6d8bb2f8e90e3f0e")
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 int64, ndim 0, dtype family int, 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 numpy as np

# Create a simple array to search within
array = np.array([1, 5, 10, 15, 20, 25, 30])

# Create a value to find the nearest element to
value = 12

# Construct the ARGS dictionary
ARGS = {
    "array": array,
    "value": value
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int64, ndim 0, dtype family int, 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
get_thresholds Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/evaluation.py
code served (permissive licence) · get_code("6bb2f76b61f0c130")
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 numpy as np
import numba

torch_seed = 0

# Analyze the function:
# - scores: needs to be a numpy array that supports .sort() and slicing
# - num_gt: integer (number of ground truth)
# - num_pr_points: integer (number of precision-recall points)

# Create small test data
scores = np.array([0.9, 0.7, 0.5, 0.3, 0.1], dtype=np.float64)
num_gt = 10
num_pr_points = 11

ARGS = {
    "scores": scores,
    "num_gt": num_gt,
    "num_pr_points": num_pr_points,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numba, numpy — 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
locate_message Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/data.py
code served (permissive licence) · get_code("aea292196d7c9068")
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 int64, ndim 0, dtype family int, 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 numpy as np

# Create test data for the locate_message function
# utimes: array of timestamps in ascending order
utimes = np.array([1.0, 2.5, 4.0, 6.5, 8.0, 10.0])

# utime: a single timestamp to locate
utime = 5.5

ARGS = {
    "utimes": utimes,
    "utime": utime,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int64, ndim 0, dtype family int, 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
maxpool2x2 Ran
ran; our contract draft was wrong, not the code
tarashakhurana/emergent-occ-forecasting/model.py
code served (permissive licence) · get_code("c5be4e15ce21fd81")
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 MaxPool2d. 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 = {
    "stride": 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 MaxPool2d — a MaxPool2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MaxPool2d 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
overall_filter Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/eval_utils.py
code served (permissive licence) · get_code("978da9312c343fda")
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 ndarray, ndim 1, dtype family float, shape (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 numpy as np

# Create a small numpy array of boxes
# boxes should be a 2D array where each row represents a box
# The function uses boxes.shape[0], so we need at least one dimension
boxes = np.array([
    [10.0, 20.0, 30.0, 40.0],
    [15.0, 25.0, 35.0, 45.0]
], dtype=np.float32)

ARGS = {
    "boxes": boxes
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (2,) — shape (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
poly_area Ran
ran, violated the contract we drafted
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/rotate_iou_cpu_eval.py
code served (permissive licence) · get_code("e5929ca970405ef7")
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 float64, ndim 0, 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 numpy as np

# poly_area computes the area of a polygon given x and y coordinates
# It uses np.dot and np.roll, expecting array-like inputs
# x and y should be 1D arrays of coordinates

x = np.array([0.0, 1.0, 1.0, 0.0])
y = np.array([0.0, 0.0, 1.0, 1.0])

ARGS = {
    "x": x,
    "y": y,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
polygon_clip Ran
ran; our contract draft was wrong, not the code
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once_eval/rotate_iou_cpu_eval.py
code served (permissive licence) · get_code("05b20b55c5918c4a")
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 2, 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.

# Define a simple polygon - a square in counter-clockwise order
subjectPolygon = [(0, 0), (4, 0), (4, 4), (0, 4)]

# Define a clipping polygon - a convex shape (triangle) in counter-clockwise order
clipPolygon = [(1, 1), (5, 1), (2, 5)]

ARGS = {
    "subjectPolygon": subjectPolygon,
    "clipPolygon": clipPolygon,
}
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 2, 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
sample Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/sampler.py
code served (permissive licence) · get_code("2f73fde7de6f64a3")
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 ndarray, ndim 3, dtype family float, shape (3, 5, 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 numpy as np
from scipy.special import fresnel

# Set random seed for reproducibility
np.random.seed(0)

# Define small but distinct dimensions for different purposes
v0 = 2.5  # scalar initial velocity
Kappa = 0.15  # scalar curvature parameter
T0 = np.array([1.0, 0.0])  # 2D tangent vector (direction)
N0 = np.array([0.0, 1.0])  # 2D normal vector (perpendicular)
tt = np.linspace(0, 1, 5)  # time points: shape (5,)
M = 3  # number of trajectories

ARGS = {
    "v0": v0,
    "Kappa": Kappa,
    "T0": T0,
    "N0": N0,
    "tt": tt,
    "M": M,
    "debug": False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (3, 5, 3) — shape (3, 5, 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
split_info_loader_helper Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/util/once_devkit/once.py
code served (permissive licence) · get_code("831d5db43f83b4be")
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 functools
import os.path as osp
import tempfile
import os

# Create a simple function to wrap with split_info_loader_helper
def sample_split_loader():
    """A sample function that returns a file path"""
    # Create a temporary file with some split data
    with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt') as f:
        f.write("sample1\n")
        f.write("sample2\n")
        f.write("sample3\n")
        temp_path = f.name
    return temp_path

# Apply the decorator
@functools.wraps(sample_split_loader)
def decorated_loader():
    return sample_split_loader()

# Test with a non-existent file path as well
def sample_split_loader_nonexistent():
    """A sample function that returns a non-existent file path"""
    return "/nonexistent/path/to/split_file.txt"

@functools.wraps(sample_split_loader_nonexistent)
def decorated_loader_nonexistent():
    return sample_split_loader_nonexistent()

ARGS = {
    "func": sample_split_loader
}
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 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
traverse Ran
ran on a synthesized input
tarashakhurana/emergent-occ-forecasting/rasterize.py
code served (permissive licence) · get_code("cff9bbe2b2841671")
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.

class MockNuScenes:
    def __init__(self):
        self.data = {
            'sample': {
                'token_0': {'next': 'token_1'},
                'token_1': {'next': 'token_2'},
                'token_2': {'next': ''},
            },
            'sample_data': {
                'data_0': {'next': 'data_1'},
                'data_1': {'next': 'data_2'},
                'data_2': {'next': ''},
            }
        }
    
    def get(self, name, token):
        return self.data[name][token]


nusc = MockNuScenes()
name = 'sample'
pointer = 'next'
length = 5
token = 'token_0'

ARGS = {
    'nusc': nusc,
    'name': name,
    'pointer': pointer,
    'length': length,
    'token': token
}
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
traverse_once Not yet run
unverified (harvested, no recorded run)
tarashakhurana/emergent-occ-forecasting/rasterize.py
code served (permissive licence) · get_code("f156020a9ce871dd")
How it ran

no invocation recorded.

Also present in other papers' code

Of 20 samples lifted from this paper's linked repositories, 20 carry a function identity and 0 do not. Of the 20 distinct functions with an identity, 7 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 7 later) and 13 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
div_up 9 other papers
6 dated earlier, 3 later
1908.09492 (2019-08)
Class-balanced Grouping and Sampling for Point Cloud 3D Object Detection
1911.10194 (2019-11)
Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation
1911.11288 (2019-11)
Autolabeling 3D Objects with Differentiable Rendering of SDF Shape Priors
and 6 more not named here
poly_area 3 other papers
1 dated earlier, 2 later
2105.07647 (2021-05)
FGR: Frustum-Aware Geometric Reasoning for Weakly Supervised 3D Vehicle Detection
2404.16493 (2024-04)
Commonsense Prototype for Outdoor Unsupervised 3D Object Detection
2406.07544 (2024-06)
Situational Awareness Matters in 3D Vision Language Reasoning
polygon_clip 3 other papers
1 dated earlier, 2 later
2105.07647 (2021-05)
FGR: Frustum-Aware Geometric Reasoning for Weakly Supervised 3D Vehicle Detection
2404.16493 (2024-04)
Commonsense Prototype for Outdoor Unsupervised 3D Object Detection
2406.07544 (2024-06)
Situational Awareness Matters in 3D Vision Language Reasoning
conv3x3 2 other papers
2 later
2302.13130 (2023-02)
Point Cloud Forecasting as a Proxy for 4D Occupancy Forecasting
2310.11239 (2023-10)
LiDAR-based 4D Occupancy Completion and Forecasting
deconv3x3 2 other papers
2 later
2302.13130 (2023-02)
Point Cloud Forecasting as a Proxy for 4D Occupancy Forecasting
2310.11239 (2023-10)
LiDAR-based 4D Occupancy Completion and Forecasting
convex_hull_intersection 1 other paper
1 later
2404.16493 (2024-04)
Commonsense Prototype for Outdoor Unsupervised 3D Object Detection
maxpool2x2 1 other paper
1 later
2310.11239 (2023-10)
LiDAR-based 4D Occupancy Completion and Forecasting

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.

For agents

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

get_harvested_code_for_paper("2210.01917")
get_code_for_paper("2210.01917")
have("2210.01917")

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

curl https://syntology.ai/api/ran/2210.01917.json

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

[![Syntology run record](https://syntology.ai/api/ran/2210.01917.svg)](https://syntology.ai/paper/2210.01917)

Connect an agent — have() is free.