SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2410.08151 · 2024

Progressive Autoregressive Video Diffusion Models

arXiv · PDF · Open in the Atlas

Code that ran

7 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 4 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 3 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/2410.08151.json · badge: /api/ran/2410.08151.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
desaixie/pa_vdm canonical (classifier label)
classifier basis: census
7 ran · 0 no recorded run
FunctionStatusWhere it lives
build_module Ran
ran; our contract draft was wrong, not the code
desaixie/pa_vdm/opensora/registry.py
code served (permissive licence) · get_code("af810cd485700e94")
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 SimpleModule. 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
from copy import deepcopy


# Create a simple stand-in Registry class
class SimpleRegistry:
    def __init__(self):
        self.registry = {}
    
    def register(self, name):
        def decorator(cls):
            self.registry[name] = cls
            return cls
        return decorator
    
    def build(self, cfg):
        cfg = deepcopy(cfg)
        module_type = cfg.pop('type')
        if module_type in self.registry:
            return self.registry[module_type](**cfg)
        raise ValueError(f"Unknown module type: {module_type}")


# Create a simple test module
class SimpleModule(nn.Module):
    def __init__(self, in_features=4, out_features=3):
        super().__init__()
        self.linear = nn.Linear(in_features, out_features)
    
    def forward(self, x):
        return self.linear(x)


# Register the module
registry = SimpleRegistry()
registry.registry['SimpleModule'] = SimpleModule

# Create test instances for different scenarios
torch.manual_seed(0)

# Case 1: Module as dict config
module_dict = {
    'type': 'SimpleModule',
    'in_features': 4,
    'out_features': 3
}

# Case 2: Module as nn.Module instance
module_instance = SimpleModule(in_features=4, out_features=3)

# Define ARGS - test with dict input case
ARGS = {
    'module': module_dict,
    'builder': registry
}
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 SimpleModule — a SimpleModule output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModule 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
calculate_psnr Ran
ran; our contract draft was wrong, not the code
desaixie/pa_vdm/eval/vae/cal_psnr.py
code served (permissive licence) · get_code("1d84e2bdaf34e46f")
How it ran

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

import torch
import numpy as np
import math
from tqdm import tqdm

torch.manual_seed(0)

# Create small test videos as torch tensors
# Shape: (num_videos, num_frames, channels, height, width)
# Using distinct sizes: 2 videos, 3 frames, 4 channels, 5 height, 6 width
videos1 = torch.randn(2, 3, 4, 5, 6)
videos2 = torch.randn(2, 3, 4, 5, 6)

# Make videos2 slightly different from videos1 to get non-infinite PSNR
videos2 = videos1 + 0.01 * torch.randn(2, 3, 4, 5, 6)

ARGS = {
    'videos1': videos1,
    'videos2': videos2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, tqdm — 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
calculate_ssim_function Ran
ran, honoured the contract we drafted
desaixie/pa_vdm/eval/vae/cal_ssim.py
code served (permissive licence) · get_code("3bdf7dc84b510bf0")
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 float64, 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 numpy as np
import cv2

# Create two small test images with the same shape
# Using 2D grayscale images (ndim == 2)
img1 = np.random.randint(0, 256, size=(32, 32), dtype=np.uint8)
img2 = np.random.randint(0, 256, size=(32, 32), dtype=np.uint8)

ARGS = {
    'img1': img1,
    'img2': img2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
img_psnr Ran
ran, honoured the contract we drafted
desaixie/pa_vdm/eval/vae/cal_psnr.py
code served (permissive licence) · get_code("5a1fb9d13411a6e9")
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 float, 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 math
import numpy as np
import torch

torch.manual_seed(0)

# img_psnr takes two image arrays and computes PSNR
# It performs: mean squared error calculation, then PSNR formula
# Both parameters should be numpy arrays or array-like with numeric values

# Create two small distinct image arrays
img1 = np.random.rand(2, 4, 3, 3).astype(np.float32) * 255
img2 = np.random.rand(2, 4, 3, 3).astype(np.float32) * 255

ARGS = {
    "img1": img1,
    "img2": img2,
}
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 float, 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
ssim Ran
ran, honoured the contract we drafted
desaixie/pa_vdm/eval/vae/cal_ssim.py
code served (permissive licence) · get_code("897aa81e3eb07ee1")
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 float64, 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 cv2
import numpy as np

# Create two small sample images (uint8 grayscale images)
# Making them reasonably sized to have enough pixels after the [5:-5, 5:-5] crop
img1 = np.random.randint(0, 256, size=(32, 32), dtype=np.uint8)
img2 = np.random.randint(0, 256, size=(32, 32), dtype=np.uint8)

ARGS = {
    "img1": img1,
    "img2": img2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
desaixie/pa_vdm/eval/vae/cal_lpips.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from showlab/moviebecnh, is pointer only (licence: NONE) · get_code("a0bb3768ee24ee05")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from showlab/moviebecnh, 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.

trans Ran
ran, honoured the contract we drafted
desaixie/pa_vdm/eval/vae/cal_flolpips.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from showlab/moviebecnh, is pointer only (licence: NONE) · get_code("4f6235b68bad4f04")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from showlab/moviebecnh, 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.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 7 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 7 later) and 0 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
img_psnr 2 other papers
2 later
2411.15262 (2024-11)
MovieBench: A Hierarchical Movie Level Dataset for Long Video Generation
2506.00329 (2025-06)
Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation
trans 2 other papers
2 later
2411.15262 (2024-11)
MovieBench: A Hierarchical Movie Level Dataset for Long Video Generation
2506.00329 (2025-06)
Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation
trans 2 other papers
2 later
2411.15262 (2024-11)
MovieBench: A Hierarchical Movie Level Dataset for Long Video Generation
2506.00329 (2025-06)
Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation
build_module 1 other paper
1 later
2503.09595 (2025-03)
PISA Experiments: Exploring Physics Post-Training for Video Diffusion Models by Watching Stuff Drop
calculate_psnr 1 other paper
1 later
2411.15262 (2024-11)
MovieBench: A Hierarchical Movie Level Dataset for Long Video Generation
calculate_ssim_function 1 other paper
1 later
2506.00329 (2025-06)
Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation
ssim 1 other paper
1 later
2506.00329 (2025-06)
Foresight: Adaptive Layer Reuse for Accelerated and High-Quality Text-to-Video Generation

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

Current frontier video diffusion models have demonstrated remarkable results at generating high-quality videos. However, they can only generate short video clips, normally around 10 seconds or 240 frames, due to computation limitations during training. Existing methods naively achieve autoregressive long video generation by directly placing the ending of the previous clip at the front of the attention window as conditioning, which leads to abrupt scene changes, unnatural motion, and error accumulation. In this work, we introduce a more natural formulation of autoregressive long video generation by revisiting the noise level assumption in video diffusion models. Our key idea is to 1. assign the frames with per-frame, progressively increasing noise levels rather than a single noise level and 2. denoise and shift the frames in small intervals rather than all at once. This allows for smoother attention correspondence among frames with adjacent noise levels, larger overlaps between the attention windows, and better propagation of information from the earlier to the later frames. Video diffusion models equipped with our progressive noise schedule can autoregressively generate long videos with much improved fidelity compared to the baselines and minimal quality degradation over time. We present the first results on text-conditioned 60-second (1440 frames) long video generation at a quality close to frontier models. Code and video results are available at https://desaixie.github.io/pa-vdm/.

For agents

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

get_harvested_code_for_paper("2410.08151")
get_code_for_paper("2410.08151")
have("2410.08151")

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

curl https://syntology.ai/api/ran/2410.08151.json

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

[![Syntology run record](https://syntology.ai/api/ran/2410.08151.svg)](https://syntology.ai/paper/2410.08151)

Connect an agent — have() is free.