SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2202.10643 · 2022

Equivariant Graph Hierarchy-Based Neural Networks

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, 0 violated it and 1 has no contract verdict; 6 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/2202.10643.json · badge: /api/ran/2202.10643.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
hanjq17/eghn canonical (classifier label)
classifier basis not recorded on the link
7 ran · 1 no recorded run
FunctionStatusWhere it lives
aggregate Ran
ran; our contract draft was wrong, not the code
hanjq17/eghn/model/basic.py
pointer only (licence: MIT); the stored copy get_code serves, from ALRhub/mango, is pointer only (licence: NONE) · get_code("c13557c4c4b42aff")
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 ALRhub/mango, 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.

collector Ran
ran; our contract draft was wrong, not the code
hanjq17/eghn/utils.py
code served (permissive licence) · get_code("d46f2eac5def9db4")
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 list. 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)

# Create sample batch elements
# Each element is a tuple of (loc, vel, edge_attr, charges, loc_end, vel_end, other_data, graph_idx)
batch_element_1 = (
    torch.randn(3, 4),      # loc: [N1=3, 4]
    torch.randn(3, 4),      # vel: [N1=3, 4]
    torch.randn(3, 2),      # edge_attr: [N1=3, 2]
    torch.randn(3, 1),      # charges: [N1=3, 1]
    torch.randn(3, 4),      # loc_end: [N1=3, 4]
    torch.randn(3, 4),      # vel_end: [N1=3, 4]
    torch.tensor([0, 1, 2]), # other_data: [N1=3]
    torch.tensor([0, 0, 0])  # graph_idx: [N1=3]
)

batch_element_2 = (
    torch.randn(5, 4),      # loc: [N2=5, 4]
    torch.randn(5, 4),      # vel: [N2=5, 4]
    torch.randn(5, 2),      # edge_attr: [N2=5, 2]
    torch.randn(5, 1),      # charges: [N2=5, 1]
    torch.randn(5, 4),      # loc_end: [N2=5, 4]
    torch.randn(5, 4),      # vel_end: [N2=5, 4]
    torch.tensor([0, 1, 2, 3, 4]), # other_data: [N2=5]
    torch.tensor([1, 1, 1, 1, 1])  # graph_idx: [N2=5]
)

batch = [batch_element_1, batch_element_2]

ARGS = {
    "batch": batch
}
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 list — a list 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
collector_simulation Ran
ran; our contract draft was wrong, not the code
hanjq17/eghn/utils.py
code served (permissive licence) · get_code("09a6af5717cfe3e4")
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)

# Create a batch of 2 samples, each with different node counts
# Each sample is a tuple of 8 elements: (loc, vel, edges, edge_attr, local_edge_mask, charges, loc_end, vel_end)

# Sample 1: 3 nodes, 2 features
loc_1 = torch.randn(3, 2)
vel_1 = torch.randn(3, 2)
edges_1 = torch.tensor([[0, 1, 2], [1, 2, 0]], dtype=torch.long)  # shape (2, 3)
edge_attr_1 = torch.randn(3, 4)
local_edge_mask_1 = torch.ones(3, 1)
charges_1 = torch.randn(3, 1)
loc_end_1 = torch.randn(3, 2)
vel_end_1 = torch.randn(3, 2)

sample_1 = (loc_1, vel_1, edges_1, edge_attr_1, local_edge_mask_1, charges_1, loc_end_1, vel_end_1)

# Sample 2: 4 nodes, 2 features
loc_2 = torch.randn(4, 2)
vel_2 = torch.randn(4, 2)
edges_2 = torch.tensor([[0, 1, 2, 3], [1, 2, 3, 0]], dtype=torch.long)  # shape (2, 4)
edge_attr_2 = torch.randn(4, 4)
local_edge_mask_2 = torch.ones(4, 1)
charges_2 = torch.randn(4, 1)
loc_end_2 = torch.randn(4, 2)
vel_end_2 = torch.randn(4, 2)

sample_2 = (loc_2, vel_2, edges_2, edge_attr_2, local_edge_mask_2, charges_2, loc_end_2, vel_end_2)

# Batch of 2 samples
batch = [sample_1, sample_2]

ARGS = {
    'batch': batch
}
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
do_padding Ran
ran on a synthesized input
hanjq17/eghn/utils.py
code served (permissive licence) · get_code("559b84475945a669")
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 (15, 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 list of tensors with different first dimensions but same second dimension
tensor_list = [
    torch.randn(3, 4),  # 3 rows, 4 columns
    torch.randn(5, 4),  # 5 rows, 4 columns
    torch.randn(2, 4),  # 2 rows, 4 columns
]

ARGS = {
    "tensor_list": tensor_list,
    "padding": True,
}
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 (15, 4) — shape (15, 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
parse_amc Ran
ran; our contract draft was wrong, not the code
hanjq17/eghn/motion/amc_parser.py
code served (permissive licence) · get_code("00b1bc1dee272f18")
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 list, 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 tempfile
import os

# Create a temporary file with valid AMC format content
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.amc')
try:
    # Write AMC format content
    amc_content = """:DEGREES
3
hip 10.0 20.0 30.0
chest 5.0 15.0
lhip 12.0 22.0 32.0
5
hip 11.0 21.0 31.0
chest 6.0 16.0
lhip 13.0 23.0 33.0
"""
    temp_file.write(amc_content)
    temp_file.close()
    file_path = temp_file.name
    
    ARGS = {
        "file_path": file_path
    }
finally:
    pass
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list 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
read_line Ran
ran; our contract draft was wrong, not the code
hanjq17/eghn/motion/amc_parser.py
code served (permissive licence) · get_code("2cafa259d5ab3508")
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)

# stream: a list of strings, where each string contains space-separated tokens
stream = [
    "hello world test",
    "foo bar",
    "one two three four",
]

# idx: an integer index into the stream
idx = 0

ARGS = {
    "stream": stream,
    "idx": idx,
}
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
unsorted_segment_mean Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
hanjq17/eghn/model/basic.py
code served (permissive licence) · get_code("bc66f19be57720e7")
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)

# Create sample data tensor with shape (batch_size, feature_dim)
# Using distinct sizes: 2 for batch, 4 for features
data = torch.randn(2, 4)

# Create segment_ids tensor that maps each data point to a segment
# Values should be in range [0, num_segments)
segment_ids = torch.tensor([0, 1])

# Number of segments to aggregate into
num_segments = 2

ARGS = {
    "data": data,
    "segment_ids": segment_ids,
    "num_segments": num_segments,
}
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
collate_mda Not yet run
unverified (harvested, no recorded run)
hanjq17/eghn/mdanalysis/dataset.py
code served (permissive licence) · get_code("7d820c2f22842d1c")
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, 7 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 7 later) and 1 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
unsorted_segment_mean 20 other papers
3 dated earlier, 17 later · 1 only as an assembly brick
1810.04805 (2018-10)
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
2102.09844 (2021-02)
E(n) Equivariant Graph Neural Networks
2110.14811 (2021-10)
SE(3) Equivariant Graph Neural Networks with Complete Local Frames
and 17 more not named here
aggregate 3 other papers
3 later
2401.11037 (2024-01)
Equivariant Graph Neural Operator for Modeling 3D Dynamics
2510.04233 (2025-10)
PAINET: A Principled Efficient Transformer for 3D Dynamics Modeling
2605.20978 (2026-05)
Point Cloud Sequence Encoding for Material-conditioned Graph Network Simulators
parse_amc 2 other papers
2 later
2308.16822 (2023-08)
Latent Variable Multi-output Gaussian Processes for Hierarchical Datasets
2401.11037 (2024-01)
Equivariant Graph Neural Operator for Modeling 3D Dynamics
read_line 2 other papers
2 later
2308.16822 (2023-08)
Latent Variable Multi-output Gaussian Processes for Hierarchical Datasets
2401.11037 (2024-01)
Equivariant Graph Neural Operator for Modeling 3D Dynamics
collate_mda 1 other paper
1 later
2401.11037 (2024-01)
Equivariant Graph Neural Operator for Modeling 3D Dynamics
collector 1 other paper
1 later
2401.11037 (2024-01)
Equivariant Graph Neural Operator for Modeling 3D Dynamics
collector_simulation 1 other paper
1 later
2401.11037 (2024-01)
Equivariant Graph Neural Operator for Modeling 3D Dynamics

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

Equivariant Graph neural Networks (EGNs) are powerful in characterizing the dynamics of multi-body physical systems. Existing EGNs conduct flat message passing, which, yet, is unable to capture the spatial/dynamical hierarchy for complex systems particularly, limiting substructure discovery and global information fusion. In this paper, we propose Equivariant Hierarchy-based Graph Networks (EGHNs) which consist of the three key components: generalized Equivariant Matrix Message Passing (EMMP) , E-Pool and E-UpPool. In particular, EMMP is able to improve the expressivity of conventional equivariant message passing, E-Pool assigns the quantities of the low-level nodes into high-level clusters, while E-UpPool leverages the high-level information to update the dynamics of the low-level nodes. As their names imply, both E-Pool and E-UpPool are guaranteed to be equivariant to meet physic symmetry. Considerable experimental evaluations verify the effectiveness of our EGHN on several applications including multi-object dynamics simulation, motion capture, and protein dynamics modeling.

For agents

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

get_harvested_code_for_paper("2202.10643")
get_code_for_paper("2202.10643")
have("2202.10643")

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

curl https://syntology.ai/api/ran/2202.10643.json

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

[![Syntology run record](https://syntology.ai/api/ran/2202.10643.svg)](https://syntology.ai/paper/2202.10643)

Connect an agent — have() is free.