SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.16739 · 2023

Active Implicit Object Reconstruction using Uncertainty-guided Next-Best-View Optimization

arXiv · PDF · Open in the Atlas

Code that ran

18 of 18 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 18 have no contract verdict; 0 failures of our instrument; 0 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/2303.16739.json · badge: /api/ran/2303.16739.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
hitsz-nrsl/activeimplicitrecon canonical (classifier label)
classifier basis not recorded on the link
18 ran · 0 no recorded run
FunctionStatusWhere it lives
Exp Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/model/pose.py
code served (permissive licence) · get_code("25fef5051aaf7b72")
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 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 torch

torch.manual_seed(0)

# Create a small (3,) axis-angle vector
r = torch.randn(3, dtype=torch.float32)

ARGS = {
    "r": r
}
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 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
biased_sigmoid Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/model/network.py
code served (permissive licence) · get_code("4e88fd00592210f0")
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 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

torch.manual_seed(0)

# The function biased_sigmoid takes two parameters:
# 1. x: a tensor (used as input to torch.sigmoid and subtraction)
# 2. init_bias: a scalar or tensor (subtracted from x)

# Create a small tensor for x
x = torch.randn(2, 4)

# Create a scalar bias
init_bias = torch.tensor(0.5)

ARGS = {
    "x": x,
    "init_bias": init_bias,
}
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 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
cal_angle_gain Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/task/nbv.py
code served (permissive licence) · get_code("2e7a18de4305e23c")
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 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)

# visited_poses should be a tensor of shape (batch, 4, 4) representing transformation matrices
# The function accesses visited_poses[:, :3, 3] to get position vectors
visited_poses = torch.randn(2, 4, 4)

# pose should be a tensor of shape (4, 4) representing a single transformation matrix
# The function accesses pose[:3, 3] to get the position vector
pose = torch.randn(4, 4)

ARGS = {
    "visited_poses": visited_poses,
    "pose": pose,
}
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 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
cal_move_cost Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/task/planner.py
code served (permissive licence) · get_code("9af245f0f997d41a")
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 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)

# Create two camera pose matrices (4x4 transformation matrices)
# The function uses [:3, 3] to extract the position (translation) component
cam_pose_start = torch.randn(4, 4)
cam_pose_set = torch.randn(4, 4)

# Ensure the position vectors are not zero to avoid division by zero
cam_pose_start[:3, 3] = torch.tensor([1.0, 2.0, 3.0])
cam_pose_set[:3, 3] = torch.tensor([4.0, 5.0, 6.0])

ARGS = {
    "cam_pose_start": cam_pose_start,
    "cam_pose_set": cam_pose_set,
}
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 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
cal_relative_pose Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/misc.py
code served (permissive licence) · get_code("370c067b86c2578a")
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, 4, 4) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 torch

torch.manual_seed(0)

# The function expects obj_pose and cam_pose to be either numpy arrays or torch tensors
# It calls .inverse() on them, so they must be square matrices (or batches of square matrices)
# It calls torch.bmm(), so they need batch dimension
# Create batch of 2 4x4 transformation matrices (common for SE(3) poses)

obj_pose_tensor = torch.eye(4).unsqueeze(0).repeat(2, 1, 1).float()
cam_pose_tensor = torch.eye(4).unsqueeze(0).repeat(2, 1, 1).float()

# Slightly perturb cam_pose so it's different
cam_pose_tensor[0, 0, 3] = 1.0
cam_pose_tensor[1, 1, 3] = 2.0

device = torch.device('cpu')

ARGS = {
    'obj_pose': obj_pose_tensor,
    'cam_pose': cam_pose_tensor,
    'device': device
}
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, 4, 4) — shape (2, 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
creat_side Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/tagdet.py
code served (permissive licence) · get_code("f71e9aebdeb63e6c")
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 (12, 4, 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

ARGS = {
    "tag_size": 10,
    "gap_size": 5,
    "gap_size_short": 2,
    "height": 3,
    "width": 4,
}
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 int, shape (12, 4, 3) — shape (12, 4, 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
create_board Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/tagdet.py
code served (permissive licence) · get_code("92e2be8435710ca2")
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 (6, 4, 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

ARGS = {
    "tag_size": 10,
    "gap_size": 2,
    "width": 3,
    "height": 2,
    "center": np.array([0.5, 1.0, 2.0])
}
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 (6, 4, 3) — shape (6, 4, 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
create_box Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/tagdet.py
code served (permissive licence) · get_code("e608466cd80dd356")
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 (16, 4, 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

# Analyze the function signature and body
# create_box takes:
# - tag_size: int (used in arithmetic)
# - gap_size: int (used in arithmetic)
# - gap_size_long: int (used in arithmetic)
# - gap_size_short: int (used in arithmetic)
# - height: int (used in range loops)
# - width: int (used in range loops)
# - center: np.ndarray with shape (3,) (used in subtraction)

# Create simple valid arguments
ARGS = {
    "tag_size": 10,
    "gap_size": 2,
    "gap_size_long": 5,
    "gap_size_short": 3,
    "height": 2,
    "width": 2,
    "center": np.zeros(3)
}
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 (16, 4, 3) — shape (16, 4, 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
eval_rays_information_gain Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/task/nbv.py
code served (permissive licence) · get_code("d683de2a6a110fe3")
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 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

torch.manual_seed(0)

# Create a small tensor of occupancy probabilities
# Shape: (batch_size=2, num_rays=4)
# Values should be in [0, 1] range since they represent occupancy probabilities
occs = torch.rand(2, 4)

ARGS = {
    "occs": occs
}
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 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
load_pose Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/misc.py
code served (permissive licence) · get_code("0c53aadae835c502")
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 pickle
import tempfile
import os

# Create a temporary file with pickled pose data
temp_dir = tempfile.mkdtemp()
pose_data = {
    'position': [0.1, 0.2, 0.3],
    'rotation': [0.0, 0.0, 0.0, 1.0],
    'timestamp': 123.456
}

pose_file = os.path.join(temp_dir, 'pose.pkl')
with open(pose_file, 'wb') as f:
    pickle.dump(pose_data, f)

ARGS = {
    'path': pose_file
}
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
make_c2w Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/model/pose.py
code served (permissive licence) · get_code("3c57ee0c5874a792")
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 2, dtype family float, shape (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)

# Create small 3-element tensors for axis-angle rotation and translation
r = torch.randn(3, dtype=torch.float32)
t = torch.randn(3, dtype=torch.float32)

ARGS = {
    "r": r,
    "t": t,
}
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 2, dtype family float, shape (4, 4) — shape (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
model_state_to_torch_pose Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/task/sensor_rosnode.py
code served (permissive licence) · get_code("7a4f4aa3906509a9")
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 2, dtype family float, shape (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
from scipy.spatial.transform import Rotation as R

torch.manual_seed(0)

class Position:
    def __init__(self, x, y, z):
        self.x = x
        self.y = y
        self.z = z

class Orientation:
    def __init__(self, x, y, z, w):
        self.x = x
        self.y = y
        self.z = z
        self.w = w

class Pose:
    def __init__(self, position, orientation):
        self.position = position
        self.orientation = orientation

class ModelState:
    def __init__(self, pose):
        self.pose = pose

position = Position(x=1.0, y=2.0, z=3.0)
orientation = Orientation(x=0.0, y=0.0, z=0.0, w=1.0)
pose = Pose(position=position, orientation=orientation)
model_state = ModelState(pose=pose)

ARGS = {
    'model_state': model_state,
    'device': 'cpu'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
scipy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (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
render_color_depth_opacity_weight Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/render.py
code served (permissive licence) · get_code("b3809e04669c0b23")
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. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Analyze the function signature and body:
# - occs: used as alphas, indexed with [:, :1] and [:, :-1], shape (batch, num_samples)
# - rgbs: used with unsqueeze(-1) and multiplied with weights, shape (batch, num_samples, 3) for RGB
# - z_vals: multiplied with weights and summed, shape (batch, num_samples)

# Create small tensors with distinct dimensions for clarity
batch_size = 2
num_samples = 4
color_channels = 3

# occs: occupancy/alpha values, shape (batch, num_samples)
occs = torch.rand(batch_size, num_samples)

# rgbs: RGB values, shape (batch, num_samples, color_channels)
rgbs = torch.rand(batch_size, num_samples, color_channels)

# z_vals: depth values along ray, shape (batch, num_samples)
z_vals = torch.linspace(0, 1, num_samples).unsqueeze(0).expand(batch_size, -1)

ARGS = {
    "occs": occs,
    "rgbs": rgbs,
    "z_vals": z_vals,
}
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 — a tuple output has no array shape of its own; none is 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
render_occ_rgb Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/render.py
code served (permissive licence) · get_code("aa7beb747a17265c")
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. 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 SimpleOccRGBModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.fc = nn.Linear(3, 4)
    
    def forward(self, xyz):
        # xyz shape: (N, 3) where N = num_rays * sample_points
        # output: (occs, rgbs) where occs is (N,) and rgbs is (N, 3)
        x = self.fc(xyz)
        occs = x[:, 0]  # (N,)
        rgbs = x[:, 1:4]  # (N, 3)
        return occs, rgbs

model = SimpleOccRGBModel()

# rays: shape (num_rays, 6) where first 3 are ray_o, last 3 are ray_d
num_rays = 2
rays = torch.randn(num_rays, 6)

# z_vals: shape (1, sample_points) - the depth samples along each ray
sample_points = 4
z_vals = torch.randn(1, sample_points)

ARGS = {
    "model": model,
    "rays": rays,
    "z_vals": z_vals,
    "sample_points": sample_points
}
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 — a tuple output has no array shape of its own; none is 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
ros2tensorpose Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/task/sensor_rosnode.py
code served (permissive licence) · get_code("89464cabc5ff68b2")
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 2, dtype family float, shape (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
from scipy.spatial.transform import Rotation as R

torch.manual_seed(0)

class MockPosition:
    def __init__(self):
        self.x = 1.0
        self.y = 2.0
        self.z = 3.0

class MockOrientation:
    def __init__(self):
        self.x = 0.0
        self.y = 0.0
        self.z = 0.0
        self.w = 1.0

class MockRosPose:
    def __init__(self):
        self.position = MockPosition()
        self.orientation = MockOrientation()

ros_pose = MockRosPose()

ARGS = {
    "ros_pose": ros_pose
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
scipy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (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
sample_points_even_in_bbox Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/geometry.py
code served (permissive licence) · get_code("0af9a3c70c63bee8")
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. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create rays tensor: shape (batch_size, 6) where first 3 are origin, last 3 are direction
# batch_size=2, 6 components (3 for origin + 3 for direction)
rays = torch.randn(2, 6)
# Normalize direction vectors to be valid ray directions
rays[:, 3:6] = torch.nn.functional.normalize(rays[:, 3:6], dim=1)

# sample_points: number of sample points along each ray
sample_points = 5

# obj_size_guess: half-size of the bounding cube
obj_size_guess = 0.5

# lindisp: whether to use linear disparity sampling
lindisp = False

# perturb: perturbation amount for jittering sample points
perturb = 0.3

ARGS = {
    "rays": rays,
    "sample_points": sample_points,
    "obj_size_guess": obj_size_guess,
    "lindisp": lindisp,
    "perturb": perturb,
}
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 — a tuple output has no array shape of its own; none is 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
sample_points_on_ray Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/utils/geometry.py
code served (permissive licence) · get_code("8a163decae0ed3d3")
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. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# rays: shape (batch_size, 6) where first 3 are ray origins, last 3 are ray directions
rays = torch.randn(2, 6)

# gt_depths: shape (batch_size,) ground truth depths for each ray
gt_depths = torch.tensor([0.5, 1.5], dtype=torch.float32)

# surf_points: number of surface points to sample
surf_points = 3

# strat_points: number of stratified points to sample
strat_points = 4

# sigma: standard deviation for sampling around ground truth depth
sigma = 0.1

# cube_half_size: half size of the cube for intersection calculation
cube_half_size = 1.0

# lindisp: whether to use linear disparity sampling
lindisp = False

# perturb: perturbation factor for z-value sampling
perturb = 0.3

ARGS = {
    "rays": rays,
    "gt_depths": gt_depths,
    "surf_points": surf_points,
    "strat_points": strat_points,
    "sigma": sigma,
    "cube_half_size": cube_half_size,
    "lindisp": lindisp,
    "perturb": perturb,
}
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 — a tuple output has no array shape of its own; none is 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
vec2skew Ran
ran on a synthesized input
hitsz-nrsl/activeimplicitrecon/active_recon/model/pose.py
code served (permissive licence) · get_code("ac8493adf8db5f3f")
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 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 torch

torch.manual_seed(0)

# The function vec2skew takes a single parameter v which should be a 3-element torch tensor
v = torch.randn(3, dtype=torch.float32)

ARGS = {
    "v": v
}
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 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

Also present in other papers' code

Of 18 samples lifted from this paper's linked repositories, 18 carry a function identity and 0 do not. Of the 18 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 0 later) and 15 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
Exp 1 other paper
1 dated earlier
2302.14363 (2023-02)
Efficient Implicit Neural Reconstruction Using LiDAR
make_c2w 1 other paper
1 dated earlier
2302.14363 (2023-02)
Efficient Implicit Neural Reconstruction Using LiDAR
vec2skew 1 other paper
1 dated earlier
2302.14363 (2023-02)
Efficient Implicit Neural Reconstruction Using LiDAR

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

Actively planning sensor views during object reconstruction is crucial for autonomous mobile robots. An effective method should be able to strike a balance between accuracy and efficiency. In this paper, we propose a seamless integration of the emerging implicit representation with the active reconstruction task. We build an implicit occupancy field as our geometry proxy. While training, the prior object bounding box is utilized as auxiliary information to generate clean and detailed reconstructions. To evaluate view uncertainty, we employ a sampling-based approach that directly extracts entropy from the reconstructed occupancy probability field as our measure of view information gain. This eliminates the need for additional uncertainty maps or learning. Unlike previous methods that compare view uncertainty within a finite set of candidates, we aim to find the next-best-view (NBV) on a continuous manifold. Leveraging the differentiability of the implicit representation, the NBV can be optimized directly by maximizing the view uncertainty using gradient descent. It significantly enhances the method's adaptability to different scenarios. Simulation and real-world experiments demonstrate that our approach effectively improves reconstruction accuracy and efficiency of view planning in active reconstruction tasks. The proposed system will open source at https://github.com/HITSZ-NRSL/ActiveImplicitRecon.git.

For agents

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

get_harvested_code_for_paper("2303.16739")
get_code_for_paper("2303.16739")
have("2303.16739")

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

curl https://syntology.ai/api/ran/2303.16739.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.16739.svg)](https://syntology.ai/paper/2303.16739)

Connect an agent — have() is free.