SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2005.08514 · 2020

Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory Prediction

arXiv · PDF · Open in the Atlas

Code that ran

7 of 8 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, 1 violated it and 5 have no contract verdict; 1 failure 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/2005.08514.json · badge: /api/ran/2005.08514.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
Majiker/STAR canonical (classifier label)
classifier basis: census
7 ran · 1 no recorded run
FunctionStatusWhere it lives
L2forTest Ran
ran on a synthesized input
Majiker/STAR/src/utils.py
code served (permissive licence) · get_code("3f5f02c4eea03dfc")
How it ran

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

import torch

torch.manual_seed(0)

# Analyze the function:
# - outputs: shape (seq_length, batch, 2) - used with norm(..., dim=2)
# - targets: same shape as outputs
# - obs_length: int - used as index obs_length - 1
# - lossMask: shape (seq_length, batch) - used with sum(dim=0) and indexing

seq_length = 5
batch_size = 2
output_dim = 4

# outputs and targets: (seq_length, batch, output_dim)
outputs = torch.randn(seq_length, batch_size, output_dim)
targets = torch.randn(seq_length, batch_size, output_dim)

# obs_length: integer, should be <= seq_length
obs_length = 2

# lossMask: (seq_length, batch) - binary mask
lossMask = torch.ones(seq_length, batch_size)

ARGS = {
    "outputs": outputs,
    "targets": targets,
    "obs_length": obs_length,
    "lossMask": lossMask,
}
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
L2forTestS Ran
ran on a synthesized input
Majiker/STAR/src/utils.py
code served (permissive licence) · get_code("2070ad3c2917ac08")
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, ndim 1, 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 to understand tensor shapes:
# - outputs: used with shape[1] (seq_length), norm over dim 3, indexing [batch, time, pedi]
# - targets: same shape as outputs for subtraction
# - obs_length: used as obs_length - 1 for slicing
# - lossMask: sum over dim 0 compared to seq_length, indexing [time, pedi]
# - num_samples: default 20, used for division

# outputs and targets need shape: [num_samples, seq_length, num_pedi, coordinate_dim]
# lossMask needs shape: [seq_length, num_pedi]

num_samples = 20
seq_length = 5
num_pedi = 3
coord_dim = 2
obs_length = 2

# Create outputs and targets with shape [num_samples, seq_length, num_pedi, coord_dim]
outputs = torch.randn(num_samples, seq_length, num_pedi, coord_dim)
targets = torch.randn(num_samples, seq_length, num_pedi, coord_dim)

# Create lossMask with shape [seq_length, num_pedi]
# Make some pedestrians "full" (all timesteps present) and some not
lossMask = torch.ones(seq_length, num_pedi)
# Mark some pedestrians as incomplete by setting some values to 0
lossMask[0, 0] = 0  # First pedestrian missing at first timestep
lossMask[1, 1] = 0  # Second pedestrian missing at second timestep
# Third pedestrian is complete (all 1s)

ARGS = {
    "outputs": outputs,
    "targets": targets,
    "obs_length": obs_length,
    "lossMask": lossMask,
    "num_samples": num_samples,
}
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, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
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
getLossMask Ran
ran on a synthesized input
Majiker/STAR/src/utils.py
code served (permissive licence) · get_code("e3ea2fe91ee7f42e")
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

torch.manual_seed(0)

# outputs: used with .dim() and .shape[0] or .shape[1]
# Should be either 3D or 2D tensor
outputs = torch.randn(5, 2, 4)  # seq_length=5, batch=2, features=4

# node_first: multiplied element-wise with seq_list, so should be same shape as seq_list rows
# seq_list has shape[1] used in lossmask initialization
node_first = torch.ones(2)  # batch_size=2

# seq_list: used with .shape[1] and indexed with [framenum]
# Should be 2D: (seq_length, batch_size)
seq_list = torch.ones(5, 2)  # seq_length=5, batch=2

# using_cuda: boolean flag
using_cuda = False

ARGS = {
    "outputs": outputs,
    "node_first": node_first,
    "seq_list": seq_list,
    "using_cuda": using_cuda
}
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
get_noise Ran
ran on a synthesized input
Majiker/STAR/src/star.py
code served (permissive licence) · get_code("bf98ee08b0615ac8")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

ARGS = {
    'shape': (2, 4),
    'noise_type': 'gaussian'
}
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
get_subsequent_mask Ran
ran, violated the contract we drafted
Majiker/STAR/src/star.py
code served (permissive licence) · get_code("1eedfb10edb08c8a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (1, 4, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a small tensor with batch size 2 and sequence length 4
seq = torch.randn(2, 4)

ARGS = {
    "seq": seq
}
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 3, dtype family float, shape (1, 4, 4) — shape (1, 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
linear Ran
ran; our contract draft was wrong, not the code
Majiker/STAR/src/multi_attention_forward.py
pointer only (licence: MIT); the stored copy get_code serves, from tsinghua-fib-lab/SPDiff, is pointer only (licence: NONE) · get_code("93e49f2486b2b26d")
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 tsinghua-fib-lab/SPDiff, 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.

multi_head_attention_forward Ran
ran on a synthesized input
Majiker/STAR/src/multi_attention_forward.py
pointer only (licence: MIT); the stored copy get_code serves, from tsinghua-fib-lab/SPDiff, is pointer only (licence: NONE) · get_code("4a5f247d4b18ad03")
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 tsinghua-fib-lab/SPDiff, 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.

load_arg Not yet run
unverified (harvested, no recorded run)
Majiker/STAR/trainval.py
code served (permissive licence) · get_code("72b27f80f70cdeb4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 4 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 3 later) and 4 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
get_subsequent_mask 6 other papers
2 dated earlier, 4 later · 1 only as an assembly brick
1706.03762 (2017-06)
Attention Is All You Need
1905.05460 (2019-05)
Cognitive Graph for Multi-Hop Reading Comprehension at Scale
aaai_20341 (2022 (year only))
ShuttleNet: Position-Aware Fusion of Rally Progress and Player Styles for Stroke Forecasting in Badminton
and 3 more not named here
L2forTest 1 other paper
1 dated earlier
1903.02793 (2019-03)
SR-LSTM: State Refinement for LSTM towards Pedestrian Trajectory Prediction
linear 1 other paper
1 later
2402.06680 (2024-02)
Social Physics Informed Diffusion Model for Crowd Simulation
multi_head_attention_forward 1 other paper
1 later
2402.06680 (2024-02)
Social Physics Informed Diffusion Model for Crowd Simulation

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

Understanding crowd motion dynamics is critical to real-world applications, e.g., surveillance systems and autonomous driving. This is challenging because it requires effectively modeling the socially aware crowd spatial interaction and complex temporal dependencies. We believe attention is the most important factor for trajectory prediction. In this paper, we present STAR, a Spatio-Temporal grAph tRansformer framework, which tackles trajectory prediction by only attention mechanisms. STAR models intra-graph crowd interaction by TGConv, a novel Transformer-based graph convolution mechanism. The inter-graph temporal dependencies are modeled by separate temporal Transformers. STAR captures complex spatio-temporal interactions by interleaving between spatial and temporal Transformers. To calibrate the temporal prediction for the long-lasting effect of disappeared pedestrians, we introduce a read-writable external memory module, consistently being updated by the temporal Transformer. We show that with only attention mechanism, STAR achieves state-of-the-art performance on 5 commonly used real-world pedestrian prediction datasets.

For agents

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

get_harvested_code_for_paper("2005.08514")
get_code_for_paper("2005.08514")
have("2005.08514")

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

curl https://syntology.ai/api/ran/2005.08514.json

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

[![Syntology run record](https://syntology.ai/api/ran/2005.08514.svg)](https://syntology.ai/paper/2005.08514)

Connect an agent — have() is free.