SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2407.03241 · ICML · 2024

Terrain Classification Enhanced with Uncertainty for Space Exploration Robots from Proprioceptive Data

Matias Valdenegro-Toro, Mariela De Lucas Álvarez, Jichen Guo, Ra Úl Domíguez, Mariela De, Lucas Álvarez, De Lucas

arXiv · PDF · Open in the Atlas

Code that ran

6 of 7 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, 0 violated it and 0 have no contract verdict; 5 failures of our instrument; 1 no recorded run), measured 2026-11-22; 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/2407.03241.json · badge: /api/ran/2407.03241.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
automl/HpBandSter canonical (classifier label)
classifier basis: census
6 ran · 1 no recorded run
FunctionStatusWhere it lives
concurrent_runs_over_time Ran
ran; our contract draft was wrong, not the code
automl/HpBandSter/hpbandster/visualization.py
code served (permissive licence) · get_code("fbf3936a157323fa")
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
import matplotlib.pyplot as plt

class MockRun:
    def __init__(self, started, finished):
        self.time_stamps = {
            'started': started,
            'finished': finished
        }

# Create sample runs with distinct time intervals
run1 = MockRun(started=0.0, finished=5.0)
run2 = MockRun(started=2.0, finished=8.0)
run3 = MockRun(started=4.0, finished=9.0)
run4 = MockRun(started=6.0, finished=12.0)

runs = [run1, run2, run3, run4]

ARGS = {
    'runs': runs,
    'num_points': 512,
    'show': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy — 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
default_tool_tips Ran
ran; our contract draft was wrong, not the code
automl/HpBandSter/hpbandster/visualization.py
code served (permissive licence) · get_code("d897dd20d286154b")
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
import numpy as np

class MockRun:
    def __init__(self, run_id, loss_value, started_time, finished_time, info_dict):
        self.loss = loss_value
        self.time_stamps = {
            'started': started_time,
            'finished': finished_time
        }
        self.info = info_dict
    
    def __getitem__(self, key):
        if key == 'time_stamps':
            return self.time_stamps
        raise KeyError(key)

class MockResultObject:
    def __init__(self):
        self.id_to_config = {}
        self.id_to_runs = {}
    
    def add_config(self, run_id, config, config_info):
        self.id_to_config[run_id] = {
            'config': config,
            'config_info': config_info
        }
    
    def add_runs(self, run_id, runs):
        self.id_to_runs[run_id] = runs
    
    def get_id2config_mapping(self):
        return self.id_to_config
    
    def get_runs_by_id(self, run_id):
        return self.id_to_runs.get(run_id, [])

result_object = MockResultObject()

run1 = MockRun(
    run_id=0,
    loss_value=0.5,
    started_time=100.0,
    finished_time=150.0,
    info_dict={'epoch': 10, 'batch': 32}
)
run2 = MockRun(
    run_id=0,
    loss_value=0.3,
    started_time=150.0,
    finished_time=200.0,
    info_dict={'epoch': 20, 'batch': 32}
)

result_object.add_config(
    run_id=0,
    config={'learning_rate': 0.001, 'batch_size': 32},
    config_info={'model_type': 'cnn', 'optimizer': 'adam'}
)
result_object.add_runs(run_id=0, runs=[run1, run2])

run3 = MockRun(
    run_id=1,
    loss_value=0.4,
    started_time=200.0,
    finished_time=250.0,
    info_dict={'epoch': 15}
)

result_object.add_config(
    run_id=1,
    config={'learning_rate': 0.0001, 'batch_size': 64},
    config_info={'model_type': 'mlp'}
)
result_object.add_runs(run_id=1, runs=[run3])

learning_curves = {
    0: [0.5, 0.4, 0.3],
    1: [0.4, 0.35]
}

ARGS = {
    'result_object': result_object,
    'learning_curves': learning_curves,
    'include_run_info': True
}
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 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
extract_HBS_learning_curves Ran
ran, honoured the contract we drafted
automl/HpBandSter/hpbandster/core/result.py
code served (permissive licence) · get_code("081fbf1e30807afe")
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 3, 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 MockRun:
    def __init__(self, budget, loss):
        self.budget = budget
        self.loss = loss

runs = [
    MockRun(budget=1, loss=0.5),
    MockRun(budget=3, loss=0.3),
    MockRun(budget=2, loss=None),
    MockRun(budget=4, loss=0.1),
]

ARGS = {
    "runs": runs
}
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 3, 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
finished_runs_over_time Ran
ran; our contract draft was wrong, not the code
automl/HpBandSter/hpbandster/visualization.py
code served (permissive licence) · get_code("a6b8b44c53fb5dee")
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
import matplotlib.pyplot as plt

class Run:
    def __init__(self, budget, finished_time):
        self.budget = budget
        self.time_stamps = {'finished': finished_time}

# Create sample runs with different budgets and finished times
run1 = Run(budget=1.0, finished_time=10.5)
run2 = Run(budget=1.0, finished_time=20.3)
run3 = Run(budget=1.0, finished_time=35.7)
run4 = Run(budget=2.0, finished_time=15.2)
run5 = Run(budget=2.0, finished_time=25.8)
run6 = Run(budget=3.0, finished_time=12.1)

runs = [run1, run2, run3, run4, run5, run6]

ARGS = {
    'runs': runs,
    'show': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy — 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
nic_name_to_host Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
automl/HpBandSter/hpbandster/core/nameserver.py
code served (permissive licence) · get_code("363fcb28d5e4fd90")
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 str. 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 numpy as np
from unittest.mock import patch, MagicMock

# The function uses netifaces.ifaddresses() to get network interface info
# We need to mock this since we can't actually access real network interfaces

# Create a mock for netifaces
mock_ifaddresses = MagicMock()
mock_AF_INET = 2  # Standard AF_INET constant value

# Mock the return value of ifaddresses to match the expected structure
# The function calls: ifaddresses(nic_name).setdefault(AF_INET, [{'addr': 'No IP addr'}])[0]['addr']
mock_result = {
    mock_AF_INET: [{'addr': '192.168.1.100'}]
}
mock_ifaddresses.return_value = mock_result

# Patch netifaces at module import level
import sys
mock_netifaces = MagicMock()
mock_netifaces.ifaddresses = mock_ifaddresses
mock_netifaces.AF_INET = mock_AF_INET
sys.modules['netifaces'] = mock_netifaces

# Now define the arguments
ARGS = {
    'nic_name': 'eth0'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
netifaces, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
smoothing Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
automl/HpBandSter/hpbandster/optimizers/config_generators/lcnet.py
code served (permissive licence) · get_code("4a133ad9110a7e2c")
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 'smoothing' takes a single parameter 'lc' which should be a list or list-like object
# of numeric values. It iterates through the list and tracks the running minimum.
# We need to provide a simple list of numbers to test this function.

lc = [5.0, 3.0, 8.0, 2.0, 7.0, 1.0, 9.0]

ARGS = {
    "lc": lc
}
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 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
nic_name_to_host Not yet run
unverified (harvested, no recorded run)
automl/HpBandSter/hpbandster/utils.py
code served (permissive licence) · get_code("2e9a7337ea281d7b")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 6 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Terrain Classification is an essential task in space exploration, where unpredictable environments are difficult to observe using only exteroceptive sensors such as vision. Implementing Neural Network classifiers can have high performance but can be deemed untrustworthy as they lack transparency, which makes them unreliable for taking high-stakes decisions during mission planning. We address this by proposing Neural Networks with Uncertainty Quantification in Terrain Classification. We enable our Neural Networks with Monte Carlo Dropout, DropConnect, and Flipout in time series-capable architectures using only proprioceptive data as input. We use Bayesian Optimization with Hyperband for efficient hyperparameter optimization to find optimal models for trustworthy terrain classification.

For agents

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

get_harvested_code_for_paper("2407.03241")
get_code_for_paper("2407.03241")
have("2407.03241")

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

curl https://syntology.ai/api/ran/2407.03241.json

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

[![Syntology run record](https://syntology.ai/api/ran/2407.03241.svg)](https://syntology.ai/paper/2407.03241)

Connect an agent — have() is free.