SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.15873 · 2023

Confronting Ambiguity in 6D Object Pose Estimation via Score-Based Diffusion on SE(3)

arXiv · PDF · Open in the Atlas

Code that ran

17 of 23 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, 0 violated it and 17 have no contract verdict; 0 failures of our instrument; 6 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2305.15873.json · badge: /api/ran/2305.15873.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
Ending2015a/liepose-diffusion canonical (classifier label)
classifier basis: census
17 ran · 6 no recorded run
FunctionStatusWhere it lives
broadcast_batch Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/exp/model.py
code served (permissive licence) · get_code("deb349af8fab7558")
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 ArrayImpl, ndim 2, dtype family int, shape (6, 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 jax.numpy as jnp
import numpy as np

# The function broadcast_batch takes:
# - x: a JAX array with shape (b, features) where b is batch size
# - bs: an integer representing desired batch size (must be divisible by b)

# Create a small input array
b = 2  # original batch size
features = 4  # feature dimension
x = jnp.arange(b * features).reshape((b, features))

# Create batch size that is divisible by b
bs = 6  # 6 is divisible by 2, so s = 3

ARGS = {
    "x": x,
    "bs": bs,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 2, dtype family int, shape (6, 4) — shape (6, 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
exclude Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/exploration.py
code served (permissive licence) · get_code("730c783afb3334fe")
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 torch

torch.manual_seed(0)

# The function exclude checks if any dictionary in 'excludes' is a subset of 'opt'
# by comparing their items. Both opt and excludes need to be dictionaries (or dict-like).

opt = {
    "learning_rate": 0.001,
    "batch_size": 32,
    "epochs": 10,
    "model": "resnet50"
}

excludes = [
    {"learning_rate": 0.001, "batch_size": 32},
    {"epochs": 5},
]

ARGS = {
    "opt": opt,
    "excludes": excludes
}
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 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
get_activ_fn Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/exp/model.py
code served (permissive licence) · get_code("37ecd1cfcb12483a")
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 custom_jvp. 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 jax
import jax.numpy as jnp

ARGS = {
    'activ_name': 'relu'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type custom_jvp — a custom_jvp output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a custom_jvp 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
get_allo_rotation Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/jnpfn.py
code served (permissive licence) · get_code("ca26b38b1b5c7857")
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 ArrayImpl, 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 jax.numpy as jnp

# Create small test tensors for the function parameters
# tran: translation vector - shape (3,) for a single object or (batch, 3) for multiple
tran = jnp.array([[1.0, 2.0, 3.0], [0.5, 1.0, 2.0]], dtype=jnp.float32)

# cam: camera ray direction - defaults to [0, 0, 1] if None
# We'll provide a non-default camera direction
cam = jnp.array([0.1, 0.2, 0.95], dtype=jnp.float32)

# eps: small epsilon value for numerical stability
eps = 1e-05

ARGS = {
    "tran": tran,
    "cam": cam,
    "eps": eps,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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_args Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/config.py
code served (permissive licence) · get_code("7eaf46160a8da8d3")
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 Namespace. 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 argparse
import sys

def get_args(exp_name='exp1', desc='Liepose'):
    parser = argparse.ArgumentParser(description=desc)
    parser.add_argument('--root_path', type=str, default=f'./logs/experiments/{exp_name}')
    parser.add_argument('--config', type=str, default=None, help='restore arguments from path')
    parser.add_argument('--config_explore', type=str, default=None, help='load explorations from file')
    parser.add_argument('dot_list', nargs=argparse.REMAINDER)
    return parser.parse_args()

# Mock sys.argv to simulate command line arguments
original_argv = sys.argv.copy()
sys.argv = ['test_script.py']

# Call the function with default parameters
result = get_args(exp_name='exp1', desc='Liepose')

# Restore original argv
sys.argv = original_argv

ARGS = {
    'exp_name': 'exp1',
    'desc': 'Liepose'
}
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 Namespace — a Namespace output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Namespace 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
get_obj_ray Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/npfn.py
code served (permissive licence) · get_code("363de575dd55271d")
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, 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

# Create a camera intrinsic matrix (3x3)
# Using a simple camera matrix: [[fx, 0, cx], [0, fy, cy], [0, 0, 1]]
cam = np.array([
    [500.0, 0.0, 320.0],
    [0.0, 500.0, 240.0],
    [0.0, 0.0, 1.0]
], dtype=np.float32)

# Create pixel coordinates (2D points)
# Shape: (2, 2) - batch of 2 pixels, each with (x, y)
pixel = np.array([
    [100.0, 150.0],
    [200.0, 250.0]
], dtype=np.float32)

ARGS = {
    'cam': cam,
    'pixel': pixel
}
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 (2, 3) — shape (2, 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
get_params_dict Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/exp/testbed.py
code served (permissive licence) · get_code("b2806fc7d29a3966")
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 torch.nn as nn

torch.manual_seed(0)

class MockState:
    def __init__(self):
        self.params = {
            'layer1': torch.randn(2, 4),
            'layer2': torch.randn(3, 5)
        }
        self.params_ema = {
            'layer1': torch.randn(2, 4),
            'layer2': torch.randn(3, 5)
        }

state = MockState()

ARGS = {
    'state': state,
    'key': 'layer1',
    'ema': 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 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
get_vars_dict Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/exp/testbed.py
code served (permissive licence) · get_code("7865a8902aaa8346")
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 torch.nn as nn

torch.manual_seed(0)

class State:
    def __init__(self):
        self.params = {
            'layer1': torch.randn(4, 3),
            'layer2': torch.randn(2, 5)
        }
        self.params_ema = {
            'layer1': torch.randn(4, 3),
            'layer2': torch.randn(2, 5)
        }
        self.batch_stats = {
            'layer1': torch.randn(4),
            'layer2': torch.randn(2)
        }

state = State()

ARGS = {
    'state': state,
    'key': 'layer1',
    'ema': 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 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
get_workdir Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/common.py
code served (permissive licence) · get_code("6eafe7c79b67fd43")
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 os

# The function get_workdir takes an optional 'default' parameter
# and returns the value of the 'WORKDIR' environment variable if it exists,
# otherwise returns the default value

# Set up environment variable for testing
os.environ['WORKDIR'] = '/test/workdir'

ARGS = {
    'default': None
}
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 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
homogeneous Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/npfn.py
code served (permissive licence) · get_code("d19a106e07192a8b")
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, 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

# Test case 1: 2D pixel coordinates (shape ends in 2)
p_2d = np.array([[10.0, 20.0], [30.0, 40.0]])

# Test case 2: 3D homogeneous coordinates (shape ends in 3)
p_3d = np.array([[10.0, 20.0, 2.0], [30.0, 40.0, 4.0]])

# Create ARGS dict with the 2D case as the primary test
ARGS = {
    "p": p_2d
}
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 (2, 3) — shape (2, 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
is_jaxarray Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/common.py
code served (permissive licence) · get_code("a960cfa7852754b9")
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 jax.numpy as jnp

# Create a JAX array to test with
jax_array = jnp.array([1.0, 2.0, 3.0, 4.0])

# The function takes one parameter: t
# It checks if t is an instance of jnp.ndarray
ARGS = {
    "t": jax_array
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — 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
is_jaxndarray Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/common.py
code served (permissive licence) · get_code("011869649e90d660")
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 jax.numpy as jnp
import numpy as np

# Create various JAX arrays for testing
jax_array_1d = jnp.array([1.0, 2.0, 3.0, 4.0])
jax_array_2d = jnp.array([[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]])
jax_array_3d = jnp.ones((2, 3, 4))
jax_array_4d = jnp.ones((2, 3, 4, 5))

# Create non-JAX array (numpy) for negative test
numpy_array = np.array([1.0, 2.0, 3.0])

ARGS = {
    "t": jax_array_2d,
    "nd": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, 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
iter_nested Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/nest.py
code served (permissive licence) · get_code("a70dd739f5183ba9")
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 generator. 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 typing import Any, Iterator, List, Optional, Tuple

torch.manual_seed(0)

# Test case 1: nested dict with tuples
data_1 = {
    'a': (1, 2, 3),
    'b': 4,
    'c': (5, (6, 7)),
    'd': {'e': 8, 'f': (9, 10)}
}

# Test case 2: simple tuple
data_2 = (1, 2, (3, 4))

# Test case 3: dict with tensors
data_3 = {
    'tensor_a': torch.tensor([1.0, 2.0]),
    'nested': (torch.tensor([3.0, 4.0]), 5)
}

# Test case 4: complex nested structure
data_4 = {
    'level1': {
        'level2': (torch.randn(2, 4), np.array([1, 2, 3])),
        'scalar': 42
    },
    'simple': (1, 2, 3)
}

ARGS = {
    'data': data_1,
    'sortkey': False
}
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 generator — a generator output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a generator 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
iter_nested_tuple Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/nest.py
code served (permissive licence) · get_code("19785438e9bde35f")
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 generator. 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 typing import Any, Callable, Iterator, List, Optional, Tuple

torch.manual_seed(0)

# Test case 1: tuple of dicts with nested values
data_a = {'x': 1, 'y': (2, 3)}
data_b = {'x': 4, 'y': (5, 6)}
data_tuple_1 = (data_a, data_b)

# Test case 2: tuple of dicts with tensors
data_c = {'a': torch.tensor([1, 2]), 'b': torch.tensor([3, 4])}
data_d = {'a': torch.tensor([5, 6]), 'b': torch.tensor([7, 8])}
data_tuple_2 = (data_c, data_d)

# Test case 3: tuple of nested tuples
data_e = (10, (20, 30))
data_f = (40, (50, 60))
data_tuple_3 = (data_e, data_f)

# Test case 4: complex nested structure with dicts containing tuples
data_g = {'x': (100, 200), 'y': 300}
data_h = {'x': (400, 500), 'y': 600}
data_tuple_4 = (data_g, data_h)

ARGS = {
    'data_tuple': data_tuple_1,
    'sortkey': False
}
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 generator — a generator output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a generator 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
normalize Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/jnpfn.py
code served (permissive licence) · get_code("edf5aff582f83d0e")
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 ArrayImpl, 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 jax.numpy as jnp
import numpy as np

# Create a small vector to normalize
# The function expects a vector or batch of vectors
# It computes the norm along the last axis
vec = jnp.array([[1.0, 2.0, 3.0, 4.0],
                  [5.0, 6.0, 7.0, 8.0]])

ARGS = {
    "vec": vec,
    "eps": 1e-06
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
normalize Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/npfn.py
code served (permissive licence) · get_code("f34ca9f5e8f625f7")
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 numpy as np

# Create a small test vector
vec = np.random.randn(2, 4)  # batch_size=2, dim=4

ARGS = {
    "vec": vec,
    "eps": 1e-06
}
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 (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
reg_loss Ran
ran on a synthesized input
Ending2015a/liepose-diffusion/src/liepose/utils/jnpfn.py
code served (permissive licence) · get_code("2e59e3f44b689980")
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 ArrayImpl, 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 jax.numpy as jnp

# Create tiny arrays for the function
ta = jnp.array([[1.0, 2.0, 3.0, 4.0],
                 [5.0, 6.0, 7.0, 8.0]])

out = jnp.array([[1.1, 2.1, 3.1, 4.1],
                  [5.1, 6.1, 7.1, 8.1]])

loss_type = 'l2'

ARGS = {
    'ta': ta,
    'out': out,
    'loss_type': loss_type
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
is_mat Not yet run
unverified (harvested, no recorded run)
Ending2015a/liepose-diffusion/src/liepose/metrics/r3so3.py
code served (permissive licence) · get_code("e2b550587a432c2a")
How it ran

no invocation recorded.

is_mat44 Not yet run
unverified (harvested, no recorded run)
Ending2015a/liepose-diffusion/src/liepose/metrics/so3.py
code served (permissive licence) · get_code("913a2031b6759489")
How it ran

no invocation recorded.

is_tan Not yet run
unverified (harvested, no recorded run)
Ending2015a/liepose-diffusion/src/liepose/metrics/r3so3.py
code served (permissive licence) · get_code("4194ece21d17b4dd")
How it ran

no invocation recorded.

is_vec Not yet run
unverified (harvested, no recorded run)
Ending2015a/liepose-diffusion/src/liepose/metrics/r3so3.py
code served (permissive licence) · get_code("2459541727ebbcce")
How it ran

no invocation recorded.

map_nested Not yet run
unverified (harvested, no recorded run)
Ending2015a/liepose-diffusion/src/liepose/utils/nest.py
code served (permissive licence) · get_code("b04418da77ee5324")
How it ran

no invocation recorded.

save_checkpoint Not yet run
unverified (harvested, no recorded run)
Ending2015a/liepose-diffusion/src/liepose/exp/testbed.py
code served (permissive licence) · get_code("94b060b77aaacebb")
How it ran

no invocation recorded.

Also present in other papers' code

Of 23 samples lifted from this paper's linked repositories, 23 carry a function identity and 0 do not. Of the 23 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 23 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

Addressing pose ambiguity in 6D object pose estimation from single RGB images presents a significant challenge, particularly due to object symmetries or occlusions. In response, we introduce a novel score-based diffusion method applied to the $SE(3)$ group, marking the first application of diffusion models to $SE(3)$ within the image domain, specifically tailored for pose estimation tasks. Extensive evaluations demonstrate the method's efficacy in handling pose ambiguity, mitigating perspective-induced ambiguity, and showcasing the robustness of our surrogate Stein score formulation on $SE(3)$. This formulation not only improves the convergence of denoising process but also enhances computational efficiency. Thus, we pioneer a promising strategy for 6D object pose estimation.

For agents

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

get_harvested_code_for_paper("2305.15873")
get_code_for_paper("2305.15873")
have("2305.15873")

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

curl https://syntology.ai/api/ran/2305.15873.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.15873.svg)](https://syntology.ai/paper/2305.15873)

Connect an agent — have() is free.