SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.13995 · ICLR · 2022

IGLU: Efficient GCN Training via Lazy Updates

Prateek Jain, S Deepak Narayanan, Aditya Sinha, Purushottam Kar, Sundararajan Sellamanickam

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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; 1 no recorded run), measured 2026-11-22; 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/2109.13995.json · badge: /api/ran/2109.13995.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
sdeepaknarayanan/iglu canonical (classifier label)
classifier basis: classifier_2025
3 ran · 1 no recorded run
FunctionStatusWhere it lives
get_sparse_adj Ran
ran; our contract draft was wrong, not the code
sdeepaknarayanan/iglu/src/inductive/train_arxiv.py
code served (permissive licence) · get_code("a060f56f96557546")
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) and returned csr_matrix. 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
from scipy.sparse import csr_matrix

torch.manual_seed(0)

# Construct A as a tuple of (indices, shape)
# A[0] should be a 2D array of edge indices (num_edges x 2)
# A[1] should be the data values for the sparse matrix
# A[2] should be the shape tuple

# Create edge indices: 3 edges
indices = np.array([[0, 1], [1, 2], [2, 0]], dtype=np.int32)

# Create data values: 3 values for the 3 edges
data = np.array([1.0, 2.0, 3.0], dtype=np.float32)

# Create shape tuple: 3x3 matrix
shape = (3, 3)

# Construct A as a tuple
A = (indices, data, shape)

ARGS = {
    "A": A
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, scipy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type csr_matrix — a csr_matrix output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a csr_matrix 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
mini_batch_exact Ran
ran; our contract draft was wrong, not the code
sdeepaknarayanan/iglu/src/inductive/train_ppi.py
code served (permissive licence) · get_code("9fb8a0022d6830ae")
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) 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 numpy as np
from scipy import sparse

# Create a small sparse adjacency matrix
# The function accesses adj_matrix[previous_nodes, :] and expects .indices attribute
# This means we need a sparse matrix format that supports fancy indexing and has .indices

# Create a sparse CSR matrix (Compressed Sparse Row)
# 6 nodes total, with some connections
row = np.array([0, 0, 1, 2, 2, 3, 4, 5])
col = np.array([1, 2, 3, 0, 4, 5, 1, 2])
data = np.array([1, 1, 1, 1, 1, 1, 1, 1])
adj_matrix = sparse.csr_matrix((data, (row, col)), shape=(6, 6))

# batch_nodes: starting nodes for the mini-batch
batch_nodes = np.array([0, 1])

# depth: number of hops to expand
depth = 2

ARGS = {
    "batch_nodes": batch_nodes,
    "adj_matrix": adj_matrix,
    "depth": depth,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, scipy — 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
normalize Ran
ran; our contract draft was wrong, not the code
sdeepaknarayanan/iglu/src/inductive/train_arxiv.py
code served (permissive licence) · get_code("fd53a8fff8553090")
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) and returned csc_matrix. 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 numpy as np
import scipy.sparse as sp

# Create a small sparse adjacency matrix
# Using a 4x4 matrix with some non-zero entries
rows = np.array([0, 0, 1, 2, 2, 3])
cols = np.array([1, 2, 0, 1, 3, 2])
data = np.array([1.0, 1.0, 1.0, 1.0, 1.0, 1.0])
adj = sp.csr_matrix((data, (rows, cols)), shape=(4, 4))

ARGS = {
    "adj": adj
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type csc_matrix — a csc_matrix output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a csc_matrix 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
convert_sparse_matrix_to_sparse_tensor Not yet run
unverified (harvested, no recorded run)
sdeepaknarayanan/iglu/src/inductive/train_arxiv.py
code served (permissive licence) · get_code("63bc427d8591c5ca")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 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.

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

Training multi-layer Graph Convolution Networks (GCN) using standard SGD techniques scales poorly as each descent step ends up updating node embeddings for a large portion of the graph. Recent attempts to remedy this sub-sample the graph that reduce compute but introduce additional variance and may offer suboptimal performance. This paper develops the IGLU method that caches intermediate computations at various GCN layers thus enabling lazy updates that significantly reduce the compute cost of descent. IGLU introduces bounded bias into the gradients but nevertheless converges to a first-order saddle point under standard assumptions such as objective smoothness. Benchmark experiments show that IGLU offers up to 1.2% better accuracy despite requiring up to 88% less compute.

For agents

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

get_harvested_code_for_paper("2109.13995")
get_code_for_paper("2109.13995")
have("2109.13995")

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

curl https://syntology.ai/api/ran/2109.13995.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.13995.svg)](https://syntology.ai/paper/2109.13995)

Connect an agent — have() is free.