SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2204.13683 · ECCV · 2022

KING: Generating Safety-Critical Driving Scenarios for Robust Imitation via Kinematics Gradients

Andreas Geiger, Kashyap Chitta, Apratim Bhattacharyya, Niklas Hanselmann, Katrin Renz

arXiv · PDF · Open in the Atlas

Code that ran

5 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 4 have no contract verdict; 0 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/2204.13683.json · badge: /api/ran/2204.13683.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
autonomousvision/king canonical (classifier label)
classifier basis not recorded on the link
5 ran · 1 no recorded run
FunctionStatusWhere it lives
crop_bev Ran
ran on a synthesized input
autonomousvision/king/driving_agents/king/aim_bev/data.py
code served (permissive licence) · get_code("3a340e3918c97e06")
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, 0, 106) 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 numpy as np

torch.manual_seed(0)

# Create a small BEV array (batch_size=2, height=256, width=256, channels=3)
# The function indexes it as array[:, start_x:..., start_y:...]
# So shape should be (batch, height, width) or (batch, channels, height, width)
# Looking at the indexing: array[:, start_x:start_x + crop_x, start_y:start_y + crop_y]
# This suggests shape is (batch, dim1, dim2)
array = torch.randn(2, 256, 256)

# Target image size (width, height)
target_im_size = (128, 96)

# Offsets dictionary
offsets = {
    'y_offset': 5,
    'x_offset_pos': 10,
    'x_offset_neg': 8
}

ARGS = {
    'array': array,
    'target_im_size': target_im_size,
    'offsets': offsets
}
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 Tensor, ndim 3, dtype family float, shape (2, 0, 106) — shape (2, 0, 106) 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
decode_pil_to_npy Ran
ran on a synthesized input
autonomousvision/king/driving_agents/king/aim_bev/training_utils.py
code served (permissive licence) · get_code("a1437734376e1830")
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 (15, 4, 5) 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

# Analyzing the function:
# - img parameter is used with img.shape[1], img.shape[2] (accessed as tuple/array-like)
# - img is used with bitwise operations: (img & 1 << bit_pos) >> bit_pos
# - This suggests img should be a numpy array with at least 3 dimensions
# - The bitwise operations work on numeric arrays

# Create a small test image array with shape (channels, width, height)
# Using distinct dimensions: 3 for first, 4 for second, 5 for third
img = np.random.randint(0, 256, size=(3, 4, 5), dtype=np.uint8)

ARGS = {
    "img": img
}
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 3, dtype family float, shape (15, 4, 5) — shape (15, 4, 5) 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_crop_offsets Ran
ran on a synthesized input
autonomousvision/king/driving_agents/king/aim_bev/training_utils.py
code served (permissive licence) · get_code("ab5e4f6eb1d465bd")
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 numpy as np

# Analyze the function:
# - angle_deg: a numeric value (degrees), converted to radians
# - target_im_size: used as an indexable container with [0] and [1] indices
#   representing width and height dimensions

ARGS = {
    'angle_deg': 45.0,
    'target_im_size': (256, 384)
}
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 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
load_crop_bev_png Ran
ran on a synthesized input
autonomousvision/king/driving_agents/king/aim_bev/data.py
code served (permissive licence) · get_code("3648889677ef997a")
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 int, shape (3, 274, 266) 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 PIL import Image
import tempfile
import os

# Create a temporary BEV image file for testing
temp_dir = tempfile.mkdtemp()
bev_image_path = os.path.join(temp_dir, "test_bev.png")

# Create a simple test BEV image (height=512, width=512, channels=3)
bev_img_array = np.random.randint(0, 256, (512, 512, 3), dtype=np.uint8)
bev_img = Image.fromarray(bev_img_array)
bev_img.save(bev_image_path)

# Define the arguments
ARGS = {
    'filename': bev_image_path,
    'do_augment': 1,
    'offsets': {
        'y_offset': 5,
        'x_offset_pos': 10,
        'x_offset_neg': 8,
    },
    'target_im_size': (256, 256),
    'bev_type': 'Encoded_cropped',
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — 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 int, shape (3, 274, 266) — shape (3, 274, 266) 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
normalize_imagenet Ran
ran, violated the contract we drafted
autonomousvision/king/driving_agents/king/transfuser/model.py
code served (permissive licence) · get_code("af36a314f4f4a4be")
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 4, dtype family float, shape (2, 3, 4, 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)

# The function expects a tensor x with shape (batch_size, channels, height, width)
# where channels=3 (RGB). It accesses x[:, 0], x[:, 1], x[:, 2] for the three channels.
# The tensor should contain normalized values in a reasonable range (e.g., [0, 1] or similar).

x = torch.randn(2, 3, 4, 4)  # batch=2, channels=3, height=4, width=4

ARGS = {
    "x": x
}
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 4, dtype family float, shape (2, 3, 4, 4) — shape (2, 3, 4, 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
encode_npy_to_pil Not yet run
unverified (harvested, no recorded run)
autonomousvision/king/driving_agents/king/aim_bev/training_utils.py
code served (permissive licence) · get_code("8130dc8508ed50c4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 1 later) and 5 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
normalize_imagenet 7 other papers
5 dated earlier, 2 later · 1 only as an assembly brick
1911.05063 (2019-11)
Kaolin: A PyTorch Library for Accelerating 3D Deep Learning Research
2007.10973 (2020-07)
Neural Mesh Flow: 3D Manifold Mesh Generation via Diffeomorphic Flows
2104.09224 (2021-04)
Multi-Modal Fusion Transformer for End-to-End Autonomous Driving
and 4 more not named here

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

Simulators offer the possibility of safe, low-cost development of self-driving systems. However, current driving simulators exhibit naïve behavior models for background traffic. Hand-tuned scenarios are typically added during simulation to induce safety-critical situations. An alternative approach is to adversarially perturb the background traffic trajectories. In this paper, we study this approach to safety-critical driving scenario generation using the CARLA simulator. We use a kinematic bicycle model as a proxy to the simulator's true dynamics and observe that gradients through this proxy model are sufficient for optimizing the background traffic trajectories. Based on this finding, we propose KING, which generates safety-critical driving scenarios with a 20% higher success rate than black-box optimization. By solving the scenarios generated by KING using a privileged rule-based expert algorithm, we obtain training data for an imitation learning policy. After fine-tuning on this new data, we show that the policy becomes better at avoiding collisions. Importantly, our generated data leads to reduced collisions on both held-out scenarios generated via KING as well as traditional hand-crafted scenarios, demonstrating improved robustness.

For agents

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

get_harvested_code_for_paper("2204.13683")
get_code_for_paper("2204.13683")
have("2204.13683")

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

curl https://syntology.ai/api/ran/2204.13683.json

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

[![Syntology run record](https://syntology.ai/api/ran/2204.13683.svg)](https://syntology.ai/paper/2204.13683)

Connect an agent — have() is free.