SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.09508 · 2024

SkateFormer: Skeletal-Temporal Transformer for Human Action Recognition

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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; 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/2403.09508.json · badge: /api/ran/2403.09508.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
KAIST-VICLab/SkateFormer canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
type_1_partition Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
KAIST-VICLab/SkateFormer/model/SkateFormer.py
code served (permissive licence) · get_code("f4f9afcbde2f1f62")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Tensor, ndim 4, dtype family float, shape (8, 3, 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 expects:
# - input: a 4D tensor with shape (B, C, T, V)
# - partition_size: a tuple/list with 2 elements [partition_size_T, partition_size_V]
# 
# Constraints from the function body:
# - T must be divisible by partition_size[0]
# - V must be divisible by partition_size[1]

B = 2
C = 4
T = 6  # divisible by partition_size[0]
V = 8  # divisible by partition_size[1]

partition_size = [3, 4]  # T // 3 = 2, V // 4 = 2

input_tensor = torch.randn(B, C, T, V)

ARGS = {
    'input': input_tensor,
    'partition_size': partition_size
}
Setup block drafted by
a language model, Claude Sonnet 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 4, dtype family float, shape (8, 3, 4, 4) — shape (8, 3, 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
type_1_reverse Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
KAIST-VICLab/SkateFormer/model/SkateFormer.py
code served (permissive licence) · get_code("769d18429dcdd3e4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Tensor, ndim 4, dtype family float, shape (2, 5, 8, 6) 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)

# Analyzing the function:
# - partitions: a tensor that will be reshaped
# - original_size: tuple (T, V)
# - partition_size: tuple with 2 elements (partition_size[0], partition_size[1])
#
# From the function body:
# B = int(partitions.shape[0] / (T * V / partition_size[0] / partition_size[1]))
# This means: partitions.shape[0] = B * (T * V / partition_size[0] / partition_size[1])
#
# Then: partitions.view(B, T // partition_size[0], V // partition_size[1], 
#                       partition_size[0], partition_size[1], -1)
# So partitions must have shape that can be reshaped to:
# (B, T // partition_size[0], V // partition_size[1], partition_size[0], partition_size[1], C)
# where C is the channel dimension (the -1 will infer this)

# Let's choose:
# T = 8, V = 6 (original_size)
# partition_size = (4, 3)
# B = 2
# C = 5 (some channel dimension)

T = 8
V = 6
partition_size = (4, 3)
B = 2
C = 5

# Calculate the required shape for partitions
# After view: (B, T // partition_size[0], V // partition_size[1], partition_size[0], partition_size[1], C)
# = (2, 8//4, 6//3, 4, 3, 5)
# = (2, 2, 2, 4, 3, 5)
# Total elements = 2 * 2 * 2 * 4 * 3 * 5 = 480

# partitions.shape[0] should equal B * (T * V / partition_size[0] / partition_size[1])
# = 2 * (8 * 6 / 4 / 3) = 2 * (48 / 12) = 2 * 4 = 8

partitions_shape_0 = B * (T * V // partition_size[0] // partition_size[1])
# = 2 * (8 * 6 // 4 // 3) = 2 * 4 = 8

# Total elements in partitions = 480
# partitions_shape_0 = 8
# So remaining dimension = 480 / 8 = 60

partitions = torch.randn(8, 60)

original_size = (T, V)

ARGS = {
    'partitions': partitions,
    'original_size': original_size,
    'partition_size': partition_size
}
Setup block drafted by
a language model, Claude Sonnet 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 4, dtype family float, shape (2, 5, 8, 6) — shape (2, 5, 8, 6) 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
type_2_partition Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
KAIST-VICLab/SkateFormer/model/SkateFormer.py
code served (permissive licence) · get_code("bc407d6d8f28b7ce")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned Tensor, ndim 4, dtype family float, shape (8, 3, 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)

# Analyzing the function:
# - input shape: (B, C, T, V)
# - partition_size: tuple/list of 2 elements [partition_size[0], partition_size[1]]
# - T must be divisible by partition_size[0]
# - V must be divisible by partition_size[1]

B = 2
C = 4
T = 6  # Must be divisible by partition_size[0]
V = 8  # Must be divisible by partition_size[1]

partition_size = [3, 4]  # T=6 divisible by 3, V=8 divisible by 4

input = torch.randn(B, C, T, V)

ARGS = {
    'input': input,
    'partition_size': partition_size
}
Setup block drafted by
a language model, Claude Sonnet 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 4, dtype family float, shape (8, 3, 4, 4) — shape (8, 3, 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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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

Skeleton-based action recognition, which classifies human actions based on the coordinates of joints and their connectivity within skeleton data, is widely utilized in various scenarios. While Graph Convolutional Networks (GCNs) have been proposed for skeleton data represented as graphs, they suffer from limited receptive fields constrained by joint connectivity. To address this limitation, recent advancements have introduced transformer-based methods. However, capturing correlations between all joints in all frames requires substantial memory resources. To alleviate this, we propose a novel approach called Skeletal-Temporal Transformer (SkateFormer) that partitions joints and frames based on different types of skeletal-temporal relation (Skate-Type) and performs skeletal-temporal self-attention (Skate-MSA) within each partition. We categorize the key skeletal-temporal relations for action recognition into a total of four distinct types. These types combine (i) two skeletal relation types based on physically neighboring and distant joints, and (ii) two temporal relation types based on neighboring and distant frames. Through this partition-specific attention strategy, our SkateFormer can selectively focus on key joints and frames crucial for action recognition in an action-adaptive manner with efficient computation. Extensive experiments on various benchmark datasets validate that our SkateFormer outperforms recent state-of-the-art methods.

For agents

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

get_harvested_code_for_paper("2403.09508")
get_code_for_paper("2403.09508")
have("2403.09508")

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

curl https://syntology.ai/api/ran/2403.09508.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.09508.svg)](https://syntology.ai/paper/2403.09508)

Connect an agent — have() is free.