SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2301.09632 · CVPR · 2023

HexPlane: A Fast Representation for Dynamic Scenes

Justin Johnson, Ang Cao

arXiv · PDF · Open in the Atlas

Code that ran

18 of 19 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; 2 honoured the contract we drafted, 1 violated it and 7 have no contract verdict; 8 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2301.09632.json · badge: /api/ran/2301.09632.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
Caoang327/HexPlane canonical (classifier label)
classifier basis not recorded on the link
18 ran · 1 no recorded run
FunctionStatusWhere it lives
GM_Resi Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/render/util/Sampling.py
code served (permissive licence) · get_code("33d5c85464c8631c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 1, dtype family float, shape (2,) 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)

# x: tensor of shape (batch, dim)
# y: tensor of shape (batch, dim)
# gamma: scalar tensor

x = torch.randn(2, 4)
y = torch.randn(2, 4)
gamma = torch.tensor(0.5)

ARGS = {
    "x": x,
    "y": y,
    "gamma": gamma,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (2,) — shape (2,) 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
GM_function Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/render/util/Sampling.py
code served (permissive licence) · get_code("15f59459c16f27f5")
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
import numpy as np

torch.manual_seed(0)

# GM_function takes two parameters: x and gamma
# Both appear to be numeric values or tensors that support:
# - exponentiation (**)
# - addition (+)
# - division (/)

# Create small test tensors
x = torch.randn(2, 4)
gamma = torch.randn(2, 4)

ARGS = {
    "x": x,
    "gamma": gamma
}
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 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
OctreeRender_trilinear_fast Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/render/render.py
code served (permissive licence) · get_code("3ce2ce46f2775459")
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. 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 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)

# Define a minimal model that satisfies the function's usage
class MinimalOctreeModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(10, 4)  # dummy layer
    
    def forward(self, rays_chunk, time_chunk, is_train=False, white_bg=True, ndc_ray=False, N_samples=-1):
        # rays_chunk shape: (batch, 6) for ray origins and directions
        # time_chunk shape: (batch,)
        batch_size = rays_chunk.shape[0]
        
        # Return (rgb_map, depth_map, alpha_map, z_val_map)
        rgb_map = torch.rand(batch_size, 3)
        depth_map = torch.rand(batch_size, 1)
        alpha_map = torch.rand(batch_size, 1)
        z_val_map = torch.rand(batch_size, 64)  # sample points
        
        return rgb_map, depth_map, alpha_map, z_val_map

# Construct ray data: batch of 2 rays, 6 dims (origin + direction)
rays = torch.rand(2, 6)

# Construct time data: batch of 2 time values
time = torch.rand(2)

# Construct model
model = MinimalOctreeModel()

ARGS = {
    'rays': rays,
    'time': time,
    'model': model,
    'chunk': 4096,
    'N_samples': -1,
    'ndc_ray': False,
    'white_bg': True,
    'is_train': False,
    'device': 'cpu'
}
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
RGBRender Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/model/HexPlane_Base.py
code served (permissive licence) · get_code("a0dc78e3351be1e1")
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, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn

torch.manual_seed(0)

# Create tiny tensors for each parameter
# xyz_sampled: coordinates in 3D space
xyz_sampled = torch.randn(2, 4)

# viewdirs: viewing directions
viewdirs = torch.randn(2, 3)

# features: RGB features to be returned
features = torch.randn(2, 5)

# time: temporal information
time = torch.randn(2, 1)

ARGS = {
    "xyz_sampled": xyz_sampled,
    "viewdirs": viewdirs,
    "features": features,
    "time": time,
}
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, 5) — shape (2, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
average_poses Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Caoang327/HexPlane/hexplane/dataloader/neural_3D_dataset_NDC.py
code served (permissive licence) · get_code("311310d09b78c64c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (3, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create sample poses with shape (N_images, 3, 4)
# N_images=2, 3x4 pose matrices (3 axes + 1 center position)
poses = np.array([
    [[1.0, 0.0, 0.0, 5.0],
     [0.0, 1.0, 0.0, 6.0],
     [0.0, 0.0, 1.0, 7.0]],
    [[0.9, 0.1, 0.0, 5.5],
     [-0.1, 0.9, 0.0, 6.5],
     [0.0, 0.0, 1.0, 7.5]]
], dtype=np.float32)

ARGS = {
    "poses": poses
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (3, 4) — shape (3, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
cal_n_samples Ran
ran; our contract draft was wrong, not the code
Caoang327/HexPlane/hexplane/render/util/Sampling.py
pointer only (licence: MIT); the stored copy get_code serves, from merlresearch/Gear-NeRF, is pointer only (licence: AGPL-3.0) · get_code("558487cf4a6668eb")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from merlresearch/Gear-NeRF, and AGPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

center_poses Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Caoang327/HexPlane/hexplane/dataloader/neural_3D_dataset_NDC.py
code served (permissive licence) · get_code("9da28ba9306c62af")
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 numpy as np

# Create sample pose data
# poses shape: (N_images, 3, 4)
# Each pose is a 3x4 camera extrinsic matrix [R | t]
N_images = 2
poses = np.array([
    [[1.0, 0.0, 0.0, 0.1],
     [0.0, 1.0, 0.0, 0.2],
     [0.0, 0.0, 1.0, 0.3]],
    [[0.9, 0.1, 0.0, 0.15],
     [-0.1, 0.9, 0.0, 0.25],
     [0.0, 0.0, 1.0, 0.35]]
], dtype=np.float32)

# Blender to OpenCV conversion matrix (4x4)
# Typical conversion between coordinate systems
blender2opencv = np.array([
    [1.0, 0.0, 0.0, 0.0],
    [0.0, -1.0, 0.0, 0.0],
    [0.0, 0.0, -1.0, 0.0],
    [0.0, 0.0, 0.0, 1.0]
], dtype=np.float32)

ARGS = {
    "poses": poses,
    "blender2opencv": blender2opencv
}
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 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
compute_dist_loss Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/render/util/Reg.py
code served (permissive licence) · get_code("cdfd7bcfcbb538bd")
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; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# weights: (..., S) - predicted weights of each sample along the ray
# svals: (..., S + 1) - normalized marching step of each sample along the ray
# We need S+1 samples in svals and S samples in weights
# Let's use batch size 2, S=4 (so weights is (..., 4) and svals is (..., 5))

batch_size = 2
num_samples = 4

# Create weights tensor of shape (batch_size, num_samples)
weights = torch.rand(batch_size, num_samples)

# Create svals tensor of shape (batch_size, num_samples + 1)
# svals should be monotonically increasing (normalized marching steps)
svals = torch.linspace(0, 1, num_samples + 1).unsqueeze(0).expand(batch_size, -1)
# Add some noise to make it distinct per batch element
svals = svals + torch.randn_like(svals) * 0.01
svals = torch.sort(svals, dim=-1)[0]  # Ensure monotonic

ARGS = {
    "weights": weights,
    "svals": svals,
}
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 — no output shape is stated: observe_signature.py could not observe it (import:OSError)
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
depth2dist Ran
ran; our contract draft was wrong, not the code
Caoang327/HexPlane/hexplane/dataloader/ray_utils.py
code served (permissive licence) · get_code("436d150f02b8247e")
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, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# z_vals: tensor with shape (..., N) where N is number of depth samples
# The function indexes z_vals[..., 1:] and z_vals[..., :-1], so we need at least 2 elements in last dim
# Let's use shape (2, 5) - batch of 2, 5 depth values
z_vals = torch.randn(2, 5)

# cos_angle: tensor that will be unsqueezed and broadcast with dists
# dists will have shape (2, 5) after concatenation
# cos_angle needs to broadcast with dists when unsqueezed(-1)
# So cos_angle should have shape (2,) to broadcast properly
cos_angle = torch.randn(2)

ARGS = {
    "z_vals": z_vals,
    "cos_angle": cos_angle,
}
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, 5) — shape (2, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
eval_sh Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Caoang327/HexPlane/hexplane/model/sh.py
pointer only (licence: MIT); the stored copy get_code serves, from pcl3dv/ov-nerf, is pointer only (licence: NONE) · get_code("db95ecdf5d4abe6b")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from pcl3dv/ov-nerf, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

eval_sh_bases Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Caoang327/HexPlane/hexplane/model/sh.py
code served (permissive licence) · get_code("caa9e209a9083a0f")
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) and returned Tensor, ndim 2, dtype family float, shape (2, 9) 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 eval_sh_bases takes:
# - deg: int, SH max degree (0-4)
# - dirs: torch.Tensor of shape (..., 3) representing unit directions

# Create a small batch of unit direction vectors
# Shape: (2, 3) - batch of 2 directions, each with 3 components (x, y, z)
dirs = torch.tensor([
    [1.0, 0.0, 0.0],
    [0.0, 1.0, 0.0]
], dtype=torch.float32)

# Normalize to ensure they are unit directions
dirs = dirs / torch.norm(dirs, dim=-1, keepdim=True)

# Test with degree 2 (a middle value)
deg = 2

ARGS = {
    "deg": deg,
    "dirs": dirs
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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, 9) — shape (2, 9) 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
ndc2dist Ran
ran; our contract draft was wrong, not the code
Caoang327/HexPlane/hexplane/dataloader/ray_utils.py
code served (permissive licence) · get_code("49f979cd2bfd19f5")
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)

# ndc_pts: used with indexing [:, 1:], [:, :-1], and subtraction
# needs shape (batch, sequence, features)
# Let's use (2, 4, 3) - batch=2, sequence=4, features=3
ndc_pts = torch.randn(2, 4, 3)

# cos_angle: used with .unsqueeze(-1)
# needs to be 1D tensor that will be unsqueezed
# Let's use shape (2,) to match batch dimension
cos_angle = torch.randn(2)

ARGS = {
    "ndc_pts": ndc_pts,
    "cos_angle": cos_angle,
}
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
normalize Ran
ran, violated the contract we drafted
Caoang327/HexPlane/hexplane/dataloader/neural_3D_dataset_NDC.py
code served (permissive licence) · get_code("aa33e7122eb26d8b")
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 1, dtype family float, shape (2,) 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 simple vector to normalize
v = np.array([3.0, 4.0])

ARGS = {
    "v": v
}
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 1, dtype family float, shape (2,) — shape (2,) 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
positional_encoding Ran
ran; our contract draft was wrong, not the code
Caoang327/HexPlane/hexplane/model/mlp.py
code served (permissive licence) · get_code("ff122aeac644ba09")
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, 24) 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 positions tensor: shape (batch=2, spatial_dim=3)
positions = torch.randn(2, 3)

# Create freqs parameter: number of frequency bands
freqs = 4

# Construct the ARGS dictionary
ARGS = {
    "positions": positions,
    "freqs": freqs,
}
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, 24) — shape (2, 24) 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
raw2alpha Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/model/HexPlane_Base.py
code served (permissive licence) · get_code("f651c6eca7673dec")
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)

# Construct valid arguments for raw2alpha
# sigma: torch.Tensor - shape (batch_size, num_samples)
# dist: torch.Tensor - shape (batch_size, num_samples)

sigma = torch.rand(2, 4)  # batch_size=2, num_samples=4
dist = torch.rand(2, 4)   # batch_size=2, num_samples=4

ARGS = {
    "sigma": sigma,
    "dist": dist,
}
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
rot_phi Ran
ran, honoured the contract we drafted
Caoang327/HexPlane/hexplane/dataloader/dnerf_dataset.py
code served (permissive licence) · get_code("c89b27166e424769")
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 numpy as np
import torch

torch.manual_seed(0)

ARGS = {
    "phi": 0.5
}
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 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
rot_theta Ran
ran, honoured the contract we drafted
Caoang327/HexPlane/hexplane/dataloader/dnerf_dataset.py
code served (permissive licence) · get_code("ed30f8260a70edf0")
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 numpy as np
import torch

torch.manual_seed(0)

# The function rot_theta takes a single parameter 'th' which should be a numeric value
# representing an angle in radians. The function uses it in np.cos(th) and np.sin(th)
# calls, so it should be a float or int.

th = np.pi / 4  # 45 degrees as a test value

ARGS = {
    "th": th
}
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 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
trans_t Ran
ran on a synthesized input
Caoang327/HexPlane/hexplane/dataloader/dnerf_dataset.py
code served (permissive licence) · get_code("16cca4c69adfd265")
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)

# The function trans_t takes a single parameter 't' which is used as a scalar value
# in a 4x4 transformation matrix. The parameter should be a numeric value (int or float).

ARGS = {
    "t": 2.5
}
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
get_ray_directions Not yet run
unverified (harvested, no recorded run)
Caoang327/HexPlane/hexplane/dataloader/ray_utils.py
code served (permissive licence) · get_code("c3d0dacfd7710891")
How it ran

no invocation recorded.

Also present in other papers' code

Of 19 samples lifted from this paper's linked repositories, 19 carry a function identity and 0 do not. Of the 19 distinct functions with an identity, 13 are also present in other papers whose code we hold (7 with at least one paper dated earlier, 3 the same month, 10 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.

FunctionAlso present inEarliest-datedOthers named
positional_encoding 17 other papers
2 dated earlier, 2 the same month, 13 later · 2 only as an assembly brick
2205.14870 (2022-05)
Compressible-composable NeRF via Rank-residual Decomposition
2212.00190 (2022-12)
Mixed Neural Voxels for Fast Multi-view Video Synthesis
2301.02238 (2023-01)
HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling
and 14 more not named here
eval_sh_bases 15 other papers
1 the same month, 14 later
2301.02238 (2023-01)
HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling
2302.01226 (2023-02)
Factor Fields: A Unified Framework for Neural Fields and Beyond
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
and 12 more not named here
eval_sh 14 other papers
14 later
2302.01226 (2023-02)
Factor Fields: A Unified Framework for Neural Fields and Beyond
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
2304.12461 (2023-04)
TensoIR: Tensorial Inverse Rendering
and 11 more not named here
normalize 10 other papers
3 dated earlier, 7 later
2008.02268 (2020-08)
NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections
2110.07342 (2021-10)
FILM: Following Instructions in Language with Modular Methods
2205.14870 (2022-05)
Compressible-composable NeRF via Rank-residual Decomposition
and 7 more not named here
average_poses 7 other papers
1 dated earlier, 6 later
2205.14870 (2022-05)
Compressible-composable NeRF via Rank-residual Decomposition
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
2305.11588 (2023-05)
Text2NeRF: Text-Driven 3D Scene Generation with Neural Radiance Fields
and 4 more not named here
center_poses 7 other papers
1 dated earlier, 6 later
2205.14870 (2022-05)
Compressible-composable NeRF via Rank-residual Decomposition
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
2305.11588 (2023-05)
Text2NeRF: Text-Driven 3D Scene Generation with Neural Radiance Fields
and 4 more not named here
depth2dist 3 other papers
3 later
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
2310.01821 (2023-10)
MIMO-NeRF: Fast Neural Rendering with Multi-input Multi-output Neural Radiance Fields
2401.00825 (2024-01)
Sharp-NeRF: Grid-based Fast Deblurring Neural Radiance Fields Using Sharpness Prior
get_ray_directions 3 other papers
3 later
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
2310.01821 (2023-10)
MIMO-NeRF: Fast Neural Rendering with Multi-input Multi-output Neural Radiance Fields
2401.00825 (2024-01)
Sharp-NeRF: Grid-based Fast Deblurring Neural Radiance Fields Using Sharpness Prior
ndc2dist 3 other papers
3 later
2303.03808 (2023-03)
Multiscale Tensor Decomposition and Rendering Equation Encoding for View Synthesis
2310.01821 (2023-10)
MIMO-NeRF: Fast Neural Rendering with Multi-input Multi-output Neural Radiance Fields
2401.00825 (2024-01)
Sharp-NeRF: Grid-based Fast Deblurring Neural Radiance Fields Using Sharpness Prior
cal_n_samples 2 other papers
1 the same month, 1 later · 1 only as an assembly brick
2301.02238 (2023-01)
HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling
2406.03723 (2024-06)
Gear-NeRF: Free-Viewpoint Rendering and Tracking with Motion-aware Spatio-Temporal Sampling
rot_phi 1 other paper
1 dated earlier
2204.00928 (2022-04)
SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image
rot_theta 1 other paper
1 dated earlier
2204.00928 (2022-04)
SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image
trans_t 1 other paper
1 dated earlier
2204.00928 (2022-04)
SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image

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

Modeling and re-rendering dynamic 3D scenes is a challenging task in 3D vision. Prior approaches build on NeRF and rely on implicit representations. This is slow since it requires many MLP evaluations, constraining real-world applications. We show that dynamic 3D scenes can be explicitly represented by six planes of learned features, leading to an elegant solution we call HexPlane. A HexPlane computes features for points in spacetime by fusing vectors extracted from each plane, which is highly efficient. Pairing a HexPlane with a tiny MLP to regress output colors and training via volume rendering gives impressive results for novel view synthesis on dynamic scenes, matching the image quality of prior work but reducing training time by more than 100×. Extensive ablations confirm our HexPlane design and show that it is robust to different feature fusion mechanisms, coordinate systems, and decoding mechanisms. HexPlane is a simple and effective solution for representing 4D volumes, and we hope they can broadly contribute to modeling spacetime for dynamic 3D scenes. 1

For agents

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

get_harvested_code_for_paper("2301.09632")
get_code_for_paper("2301.09632")
have("2301.09632")

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

curl https://syntology.ai/api/ran/2301.09632.json

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

[![Syntology run record](https://syntology.ai/api/ran/2301.09632.svg)](https://syntology.ai/paper/2301.09632)

Connect an agent — have() is free.