SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2412.09608 · 2024

Representing Long Volumetric Video with Temporal Gaussian Hierarchy

arXiv · PDF · Open in the Atlas

Code that ran

11 of 16 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 11 failures of our instrument; 5 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/2412.09608.json · badge: /api/ran/2412.09608.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
dendenxu/fast-gaussian-rasterization canonical (classifier label)
classifier basis: census
11 ran · 5 no recorded run
FunctionStatusWhere it lives
eval_shfs_4d_00 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/sh_utils.py
code served (permissive licence) · get_code("1787a71af287b1ef")
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; 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)

# Create the arguments for eval_shfs_4d_00
# The function expects:
# - sh: a tensor with shape (..., 1) since it indexes sh[..., 0]
# - dirs: a tensor (used in function signature but not in body)
# - dirs_t: a tensor (used in function signature but not in body)
# - l: a tensor (used in function signature but not in body)

sh = torch.randn(2, 4, 1)  # batch=2, dim=4, last_dim=1
dirs = torch.randn(2, 4)   # batch=2, dim=4
dirs_t = torch.randn(2, 4) # batch=2, dim=4
l = torch.randn(2, 4)      # batch=2, dim=4

ARGS = {
    "sh": sh,
    "dirs": dirs,
    "dirs_t": dirs_t,
    "l": l
}
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 — 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
eval_shfs_4d_10 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/sh_utils.py
code served (permissive licence) · get_code("39fe33843594b6cf")
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; 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)

# Analyze the function:
# - sh: indexed with [..., 0], [..., 1], [..., 2], [..., 3] -> needs last dim >= 4
# - dirs: indexed with [..., 0:1], [..., 1:2], [..., 2:3] -> needs last dim >= 3
# - dirs_t: parameter passed but not used in function body
# - l: parameter passed but not used in function body

# Create small tensors with batch size 2
batch_size = 2

# sh needs shape (..., 4) for the indexing [0], [1], [2], [3]
sh = torch.randn(batch_size, 4)

# dirs needs shape (..., 3) for the indexing [0:1], [1:2], [2:3]
dirs = torch.randn(batch_size, 3)

# dirs_t - not used in function, but required parameter. Keep it small.
dirs_t = torch.randn(batch_size, 3)

# l - not used in function, but required parameter. Keep it small.
l = torch.randn(batch_size, 1)

ARGS = {
    "sh": sh,
    "dirs": dirs,
    "dirs_t": dirs_t,
    "l": l,
}
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 — 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
eval_shfs_4d_20 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/sh_utils.py
code served (permissive licence) · get_code("cc01764c972d1217")
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; 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)

# Analyze the function:
# - sh: indexed with sh[..., 0] through sh[..., 8], so needs at least 9 channels in last dim
# - dirs: indexed with dirs[..., 0:1], dirs[..., 1:2], dirs[..., 2:3], so needs 3 in last dim
# - dirs_t: parameter but never used in function body
# - l: parameter but never used in function body

# Create small tensors with distinct dimensions for different meanings
batch_size = 2
num_sh_coeffs = 9
spatial_dims = 3

sh = torch.randn(batch_size, num_sh_coeffs)
dirs = torch.randn(batch_size, spatial_dims)
dirs_t = torch.randn(batch_size, spatial_dims)
l = torch.randn(batch_size, 1)

ARGS = {
    "sh": sh,
    "dirs": dirs,
    "dirs_t": dirs_t,
    "l": l,
}
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 — 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
line Ran
ran; our contract draft was wrong, not the code
dendenxu/fast-gaussian-rasterization/fast_gauss/console_utils.py
code served (permissive licence) · get_code("9da3d705ee4df35f")
How it ran

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

import torch
import re
from rich.pretty import Pretty, pretty_repr, pprint

torch.manual_seed(0)

# The function 'line' takes a single parameter 'obj' of any type
# It calls pretty_repr(obj, indent_size=0) on it
# So we need to provide a representative object that can be pretty-printed

# Create a diverse object that will produce multi-line output to test the formatting
test_obj = {
    'tensor': torch.randn(2, 3),
    'list': [1, 2, 3, 4, 5],
    'nested': {
        'a': 1,
        'b': 2,
        'c': [1, 2, 3]
    },
    'tuple': (1, 2, 3),
}

ARGS = {
    'obj': test_obj
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
rich, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
normalize Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/math_utils.py
code served (permissive licence) · get_code("4efe614e756cc631")
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; 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)

# Create a small tensor with shape (batch=2, dim=4)
x = torch.randn(2, 4)

# eps is a float with default value 1e-08
eps = 1e-08

ARGS = {
    "x": x,
    "eps": eps,
}
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 — 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
normalize_sum Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/math_utils.py
code served (permissive licence) · get_code("2495f2bb0cc57462")
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; 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)

# The function normalize_sum takes:
# - x: a torch.Tensor
# - eps: a float (default 1e-08)
#
# The function divides x by the sum along the last dimension (with keepdim=True) plus eps.
# We need a small tensor with a last dimension to sum over.

x = torch.randn(2, 4)  # batch of 2, last dimension of 4
eps = 1e-08

ARGS = {
    "x": x,
    "eps": eps,
}
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 — 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
path Ran
ran; our contract draft was wrong, not the code
dendenxu/fast-gaussian-rasterization/fast_gauss/console_utils.py
code served (permissive licence) · get_code("8c1aa0c6fe0ecaa2")
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 Text. 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 tempfile
from pathlib import Path

# Create a temporary file that exists
temp_dir = tempfile.mkdtemp()
existing_path = Path(temp_dir) / "test_file.txt"
existing_path.write_text("test content")

# Create a path string that doesn't exist
non_existing_path = str(Path(temp_dir) / "non_existent_file.txt")

ARGS = {
    "string": str(existing_path)
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
rich — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Text — a Text 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
red Ran
ran; our contract draft was wrong, not the code
dendenxu/fast-gaussian-rasterization/fast_gauss/console_utils.py
code served (permissive licence) · get_code("0a41221621cd471c")
How it ran

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

import torch

def red(string: str) -> str:
    return f'[red bold]{string}[/]'

ARGS = {
    "string": "Hello World"
}
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 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
rgb2sh0 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/gaussian_utils.py
code served (permissive licence) · get_code("a39fb48b94ddfa97")
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; 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)

# The function takes an RGB tensor and converts it to spherical harmonics coefficient 0
# It expects a tensor of RGB values (typically in range [0, 1] or similar)
# The tensor should have shape compatible with element-wise operations

rgb_tensor = torch.randn(2, 4)

ARGS = {
    "rgb": rgb_tensor
}
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 — 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
sh02rgb Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
dendenxu/fast-gaussian-rasterization/fast_gauss/gaussian_utils.py
code served (permissive licence) · get_code("f29f26764ed7c042")
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; 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)

# The function sh02rgb takes a tensor sh as input
# It performs a simple linear transformation: sh * C0 + 0.5
# So we need to provide a tensor with any shape

sh_tensor = torch.randn(2, 4)

ARGS = {
    "sh": sh_tensor
}
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 — 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
unregister_memory Ran
ran; our contract draft was wrong, not the code
dendenxu/fast-gaussian-rasterization/fast_gauss/cuda_utils.py
code served (permissive licence) · get_code("167e33b8202e20dc")
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 takes a tensor and creates an unpinned copy of it
# We need to pass a tensor as the argument
x = torch.randn(2, 4)

ARGS = {
    "x": x
}
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
FORMAT_CUDART_ERROR Not yet run
unverified (harvested, no recorded run)
dendenxu/fast-gaussian-rasterization/fast_gauss/cuda_utils.py
code served (permissive licence) · get_code("0f8fe090d005efe8")
How it ran

no invocation recorded.

load_shader_source Not yet run
unverified (harvested, no recorded run)
dendenxu/fast-gaussian-rasterization/fast_gauss/gl_utils.py
code served (permissive licence) · get_code("e70037d52fdf2b4f")
How it ran

no invocation recorded.

point_padding Not yet run
unverified (harvested, no recorded run)
dendenxu/fast-gaussian-rasterization/fast_gauss/math_utils.py
code served (permissive licence) · get_code("5f771c4bdef631b9")
How it ran

no invocation recorded.

register_memory Not yet run
unverified (harvested, no recorded run)
dendenxu/fast-gaussian-rasterization/fast_gauss/cuda_utils.py
code served (permissive licence) · get_code("672c292014882d13")
How it ran

no invocation recorded.

to_cuda Not yet run
unverified (harvested, no recorded run)
dendenxu/fast-gaussian-rasterization/fast_gauss/data_utils.py
code served (permissive licence) · get_code("9b018a4c2d6e5fb1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 16 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 16 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

This paper aims to address the challenge of reconstructing long volumetric videos from multi-view RGB videos. Recent dynamic view synthesis methods leverage powerful 4D representations, like feature grids or point cloud sequences, to achieve high-quality rendering results. However, they are typically limited to short (1~2s) video clips and often suffer from large memory footprints when dealing with longer videos. To solve this issue, we propose a novel 4D representation, named Temporal Gaussian Hierarchy, to compactly model long volumetric videos. Our key observation is that there are generally various degrees of temporal redundancy in dynamic scenes, which consist of areas changing at different speeds. Motivated by this, our approach builds a multi-level hierarchy of 4D Gaussian primitives, where each level separately describes scene regions with different degrees of content change, and adaptively shares Gaussian primitives to represent unchanged scene content over different temporal segments, thus effectively reducing the number of Gaussian primitives. In addition, the tree-like structure of the Gaussian hierarchy allows us to efficiently represent the scene at a particular moment with a subset of Gaussian primitives, leading to nearly constant GPU memory usage during the training or rendering regardless of the video length. Extensive experimental results demonstrate the superiority of our method over alternative methods in terms of training cost, rendering speed, and storage usage. To our knowledge, this work is the first approach capable of efficiently handling minutes of volumetric video data while maintaining state-of-the-art rendering quality. Our project page is available at: https://zju3dv.github.io/longvolcap.

For agents

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

get_harvested_code_for_paper("2412.09608")
get_code_for_paper("2412.09608")
have("2412.09608")

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

curl https://syntology.ai/api/ran/2412.09608.json

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

[![Syntology run record](https://syntology.ai/api/ran/2412.09608.svg)](https://syntology.ai/paper/2412.09608)

Connect an agent — have() is free.