SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1806.06498 · 2018

Conditional Affordance Learning for Driving in Urban Environments

arXiv · PDF · Open in the Atlas

Code that ran

12 of 16 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 11 have no contract verdict; 0 failures of our instrument; 4 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/1806.06498.json · badge: /api/ran/1806.06498.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
xl-sr/CAL canonical (classifier label)
classifier basis: census
12 ran · 4 no recorded run
FunctionStatusWhere it lives
calculate_scores Ran
ran on a synthesized input
xl-sr/CAL/training/metrics.py
code served (permissive licence) · get_code("e077d404c9974a5b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, 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
from sklearn.metrics import confusion_matrix

def get_intersection_union_per_class(confusion_matrix):
    number_of_labels = confusion_matrix.shape[0]
    matrix_diagonal = [confusion_matrix[i][i] for i in range(number_of_labels)]
    errors_summed_by_row = [0] * number_of_labels
    for row in range(number_of_labels):
        for column in range(number_of_labels):
            if row != column:
                errors_summed_by_row[row] += confusion_matrix[row][column]
    errors_summed_by_column = [0] * number_of_labels
    for column in range(number_of_labels):
        for row in range(number_of_labels):
            if row != column:
                errors_summed_by_column[column] += confusion_matrix[row][column]
    divisor = [0] * number_of_labels
    for i in range(number_of_labels):
        divisor[i] = matrix_diagonal[i] + errors_summed_by_row[i] + errors_summed_by_column[i]
        if matrix_diagonal[i] == 0:
            divisor[i] = 1
    return [float(matrix_diagonal[i]) / divisor[i] for i in range(number_of_labels)]

def calculate_scores(cl, pred_cl):
    cm = confusion_matrix(cl, pred_cl)
    precision, recall = ([], [])
    for i in range(1, len(cm)):
        p1 = 100 * (cm[i, i] / np.sum(cm[:, i]))
        r1 = 100 * (cm[i, i] / np.sum(cm[i, :]))
        precision.append('{:.2f}'.format(p1))
        recall.append('{:.2f}'.format(r1))
    val_acc = 100 * float(cm.trace()) / np.sum(cm)
    IoUs = get_intersection_union_per_class(cm)
    IoU_mean = 100 * np.mean(IoUs)
    return (val_acc, IoU_mean)

# Create sample ground truth labels and predictions
# Using labels 0, 1, 2, 3 for a 4-class classification problem
np.random.seed(0)
true_labels = np.array([0, 1, 2, 3, 0, 1, 2, 3, 0, 1])
predicted_labels = np.array([0, 1, 2, 3, 0, 2, 2, 3, 1, 1])

ARGS = {
    'cl': true_labels,
    'pred_cl': predicted_labels
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple 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
get_intersection_union_per_class Ran
ran on a synthesized input
xl-sr/CAL/training/metrics.py
code served (permissive licence) · get_code("b9527609161b21dc")
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

# The function takes a confusion_matrix as input
# Looking at the function body, it accesses:
# - confusion_matrix.shape[0] to get number of labels
# - confusion_matrix[i][i] for diagonal elements
# - confusion_matrix[row][column] for arbitrary indexing
# So we need a 2D array-like object (numpy array or list of lists)

# Create a small confusion matrix (4x4)
# This represents predictions vs ground truth for 4 classes
confusion_matrix_array = np.array([
    [10, 2, 1, 0],   # Class 0: 10 correct, 2 misclassified as 1, 1 as 2, 0 as 3
    [1, 8, 2, 1],    # Class 1: 1 misclassified as 0, 8 correct, 2 as 2, 1 as 3
    [0, 1, 12, 2],   # Class 2: 0 misclassified as 0, 1 as 1, 12 correct, 2 as 3
    [1, 0, 1, 9]     # Class 3: 1 misclassified as 0, 0 as 1, 1 as 2, 9 correct
], dtype=np.int64)

ARGS = {
    "confusion_matrix": confusion_matrix_array
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn — 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
get_params_from_txt Ran
ran on a synthesized input
xl-sr/CAL/python_client/agents/CAL_agent/CAL_agent.py
code served (permissive licence) · get_code("f7adbf77ff7462a2")
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 (3, 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
import tempfile
import os

# Create a temporary CSV file with numeric data
temp_dir = tempfile.gettempdir()
temp_file = os.path.join(temp_dir, "test_params.csv")

# Write some sample numeric data to the file
data = np.array([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0],
    [7.0, 8.0, 9.0]
])
np.savetxt(temp_file, data, delimiter=',')

ARGS = {
    "path": temp_file
}
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 2, dtype family float, shape (3, 3) — shape (3, 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
labels2classes Ran
ran on a synthesized input
xl-sr/CAL/training/metrics.py
code served (permissive licence) · get_code("0cf4b85580fb54e2")
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 int, shape (2, 1) 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

# The function labels2classes takes predictions (a numpy array)
# and applies argmax along axis 1, then reshapes to (-1, 1)
# We need a 2D numpy array where axis 1 represents class probabilities

# Create a small batch of predictions (batch_size=2, num_classes=4)
predictions = np.array([
    [0.1, 0.2, 0.6, 0.1],  # batch 0: argmax=2
    [0.3, 0.5, 0.1, 0.1]   # batch 1: argmax=1
])

ARGS = {
    'predictions': predictions
}
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 2, dtype family int, shape (2, 1) — shape (2, 1) 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
load_json Ran
ran on a synthesized input
xl-sr/CAL/python_client/agents/CAL_agent/perception/cal_network.py
code served (permissive licence) · get_code("45fe6ae45321b7ad")
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 json
import tempfile
import os

# Create a temporary directory and a test JSON file
temp_dir = tempfile.mkdtemp()
test_json_path = os.path.join(temp_dir, 'test_data')

# Write a valid JSON file (without the .json extension in the path)
test_data = {
    'key1': 'value1',
    'key2': [1, 2, 3],
    'key3': {'nested': 'object'}
}

with open(test_json_path + '.json', 'w') as f:
    json.dump(test_data, f)

# Define ARGS with the path parameter
ARGS = {
    'path': test_json_path
}
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 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
onehot Ran
ran on a synthesized input
xl-sr/CAL/training/dataloader.py
code served (permissive licence) · get_code("d78dd6853a27485d")
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 (3, 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 numpy as np

# Test with integer possible values
vals = np.array([1, 2, 3])
possible_vals = [1, 2, 3, 4]

ARGS = {
    'vals': vals,
    'possible_vals': possible_vals
}
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 2, dtype family float, shape (3, 4) — shape (3, 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
roll Ran
ran on a synthesized input
xl-sr/CAL/training/dataloader.py
code served (permissive licence) · get_code("f5b8b09a8822c2de")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 1, dtype family float, shape (4,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function roll(x, n) takes a tensor x and an integer n
# It concatenates the last n elements with all but the last n elements
# So x should be a 1D tensor and n should be an integer

x = torch.randn(4)  # Small 1D tensor
n = 2               # Integer for how many elements to roll

ARGS = {
    "x": x,
    "n": n
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (4,) — shape (4,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
softmax Ran
ran, violated the contract we drafted
xl-sr/CAL/python_client/agents/CAL_agent/perception/cal_network.py
pointer only (licence: MIT); the stored copy get_code serves, from jennhu/lm-pragmatics, is pointer only (licence: NONE) · get_code("9c4be464cbe63fba")
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 jennhu/lm-pragmatics, 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.

statistics_match Ran
ran on a synthesized input
xl-sr/CAL/training/make_validationset.py
code served (permissive licence) · get_code("c29bb767c0144b24")
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 bool, 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 numpy as np

# Create two dataframes with matching statistics
# The function checks discrete columns (red_light, hazard_stop, speed_sign) 
# using relative counts and continuous columns (center_distance, relative_angle, veh_distance)
# using means

np.random.seed(0)

# Create df1 with discrete and continuous columns
df1 = {
    'red_light': np.array([0, 1, 0, 1, 0, 1]),
    'hazard_stop': np.array([0, 0, 1, 1, 0, 0]),
    'speed_sign': np.array([1, 1, 1, 0, 0, 0]),
    'center_distance': np.array([0.5, 0.6, 0.55, 0.58, 0.52, 0.59]),
    'relative_angle': np.array([0.1, 0.12, 0.11, 0.13, 0.10, 0.12]),
    'veh_distance': np.array([5.0, 5.2, 4.8, 5.1, 4.9, 5.0])
}

# Create df2 with very similar statistics (within epsilon bounds)
df2 = {
    'red_light': np.array([0, 1, 0, 1, 0, 1]),  # Same distribution
    'hazard_stop': np.array([0, 0, 1, 1, 0, 0]),  # Same distribution
    'speed_sign': np.array([1, 1, 1, 0, 0, 0]),  # Same distribution
    'center_distance': np.array([0.51, 0.61, 0.54, 0.57, 0.53, 0.58]),  # Mean ~0.557 (within 0.02)
    'relative_angle': np.array([0.105, 0.115, 0.115, 0.125, 0.105, 0.115]),  # Mean ~0.113 (within 0.01)
    'veh_distance': np.array([5.1, 5.0, 4.9, 5.2, 4.8, 5.0])  # Mean ~5.0 (within 0.6)
}

ARGS = {
    'df1': df1,
    'df2': df2
}
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 bool, 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
string_to_floats Ran
ran on a synthesized input
xl-sr/CAL/python_client/agents/CAL_agent/plans/centerlines.py
code served (permissive licence) · get_code("3b7f3991efc831ea")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, 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 torch

torch.manual_seed(0)

ARGS = {
    "string": "1.5,2.3,3.7"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple 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
tile_array Ran
ran on a synthesized input
xl-sr/CAL/python_client/agents/CAL_agent/perception/net.py
code served (permissive licence) · get_code("a1051d33b7d689f7")
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 (6, 12) 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 numpy.lib.stride_tricks import as_strided

# Create a small 2D numpy array with distinct dimensions
a = np.arange(12).reshape(3, 4).astype(np.float32)

# Create small integer tile factors
b0 = 2
b1 = 3

ARGS = {
    'a': a,
    'b0': b0,
    'b1': b1,
}
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 2, dtype family float, shape (6, 12) — shape (6, 12) 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
to_np Ran
ran on a synthesized input
xl-sr/CAL/python_client/agents/CAL_agent/perception/cal_network.py
code served (permissive licence) · get_code("969c81fc5c65c88d")
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 (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
import numpy as np

torch.manual_seed(0)

# The function to_np takes a parameter 't' and accesses:
# - t.data (attribute)
# - t.data.cpu() (method call)
# Then converts it to a numpy array

# Create a small torch tensor to pass as 't'
tensor_input = torch.randn(2, 4)

ARGS = {
    "t": tensor_input
}
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 ndarray, 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
get_bool_vec Not yet run
unverified (harvested, no recorded run)
xl-sr/CAL/python_client/agents/CAL_agent/perception/net.py
code served (permissive licence) · get_code("55ad330f3d217cb7")
How it ran

no invocation recorded.

get_model Not yet run
unverified (harvested, no recorded run)
xl-sr/CAL/python_client/agents/CAL_agent/perception/net.py
code served (permissive licence) · get_code("60fc72096a074104")
How it ran

no invocation recorded.

get_model Not yet run
unverified (harvested, no recorded run)
xl-sr/CAL/training/net.py
code served (permissive licence) · get_code("0a1f3e5ee653e40d")
How it ran

no invocation recorded.

get_val_idcs Not yet run
unverified (harvested, no recorded run)
xl-sr/CAL/training/make_validationset.py
code served (permissive licence) · get_code("90be3766ee3dd2ac")
How it ran

no invocation recorded.

Also present in other papers' code

Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 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
softmax 2 other papers
2 later
2212.06801 (2022-12)
A fine-grained comparison of pragmatic language understanding in humans and language models
2404.19509 (2024-04)
Do Large Language Models Understand Conversational Implicature -- A case study with a chinese sitcom
string_to_floats 1 other paper
1 later
1911.04175 (2019-11)
Multi-Agent Connected Autonomous Driving using Deep Reinforcement Learning

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

Most existing approaches to autonomous driving fall into one of two categories: modular pipelines, that build an extensive model of the environment, and imitation learning approaches, that map images directly to control outputs. A recently proposed third paradigm, direct perception, aims to combine the advantages of both by using a neural network to learn appropriate low-dimensional intermediate representations. However, existing direct perception approaches are restricted to simple highway situations, lacking the ability to navigate intersections, stop at traffic lights or respect speed limits. In this work, we propose a direct perception approach which maps video input to intermediate representations suitable for autonomous navigation in complex urban environments given high-level directional inputs. Compared to state-of-the-art reinforcement and conditional imitation learning approaches, we achieve an improvement of up to 68 % in goal-directed navigation on the challenging CARLA simulation benchmark. In addition, our approach is the first to handle traffic lights and speed signs by using image-level labels only, as well as smooth car-following, resulting in a significant reduction of traffic accidents in simulation.

For agents

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

get_harvested_code_for_paper("1806.06498")
get_code_for_paper("1806.06498")
have("1806.06498")

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

curl https://syntology.ai/api/ran/1806.06498.json

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

[![Syntology run record](https://syntology.ai/api/ran/1806.06498.svg)](https://syntology.ai/paper/1806.06498)

Connect an agent — have() is free.