SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.14987 · ACL · 2022

Multilingual Knowledge Graph Completion with Self-Supervised Adaptive Graph Alignment

Yizhou Sun, Wei Wang, Haoming Jiang, Zijie Huang, Karthik Subbian, Zheng Li, Bing Yin, Hanqing Lu, Tianyu Cao, Steven Jobs

arXiv · PDF · Open in the Atlas

Code that ran

12 of 13 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 12 have no contract verdict; 0 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/2203.14987.json · badge: /api/ran/2203.14987.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
amzn/ss-aga-kgc canonical (classifier label)
classifier basis not recorded on the link
12 ran · 1 no recorded run
FunctionStatusWhere it lives
KNN_distance_batch Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/ssaga_model.py
code served (permissive licence) · get_code("fe7775148069f8ed")
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, 3) 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 query embedding with shape [N1, d] = [2, 4]
query_embedding = torch.randn(2, 4)

# Create candidate embedding with shape [N2, d] = [3, 4]
candidate_embedding = torch.randn(3, 4)

ARGS = {
    "query_embedding": query_embedding,
    "candidate_embedding": candidate_embedding,
}
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, 3) — shape (2, 3) 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
Ranking_all_batch Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/utils.py
code served (permissive licence) · get_code("505a2037a95f229d")
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 3, 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)

# predicted_t: a batch of predicted embeddings
# Shape should be (batch_size, embedding_dim)
batch_size = 2
embedding_dim = 4
predicted_t = torch.randn(batch_size, embedding_dim)

# embedding_matrix: matrix of all candidate embeddings
# Shape should be (num_entities, embedding_dim)
num_entities = 3
embedding_matrix = torch.randn(num_entities, embedding_dim)

# k: number of top candidates to return (or None for all)
k = 2

ARGS = {
    "predicted_t": predicted_t,
    "embedding_matrix": embedding_matrix,
    "k": k,
}
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 3, 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
extract_entities Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/validate.py
code served (permissive licence) · get_code("3689ab93616e1197")
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 int; 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
import numpy as np

torch.manual_seed(0)

# The function extract_entities takes id_score_tuples and extracts just the entity IDs
# id_score_tuples should be an iterable of (id, score) pairs
# The function iterates through and unpacks each tuple, keeping only the first element

id_score_tuples = [
    (101, 0.95),
    (202, 0.87),
    (303, 0.76),
    (404, 0.65)
]

ARGS = {
    "id_score_tuples": id_score_tuples
}
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 list, ndim 1, dtype family int — 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
filt_hits_at_n Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/validate.py
code served (permissive licence) · get_code("e390c35ce245a1b7")
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 logging
import pandas as pd

logging.basicConfig(level=logging.INFO)

# Create a small results dataframe
results = pd.DataFrame({
    'h': [1, 2, 3],
    'r': [10, 11, 12],
    't': [100, 101, 102],
    'en': [
        [(100, 0.9), (200, 0.8), (300, 0.7)],
        [(101, 0.95), (201, 0.85)],
        [(102, 0.92), (202, 0.82), (302, 0.72)]
    ]
})

# Create hr2t_train mapping - (h,r) -> set of valid target entities
hr2t_train = {
    (1, 10): {100},
    (2, 11): {101, 201},
}

ARGS = {
    'results': results,
    'lang': 'en',
    'hr2t_train': hr2t_train,
    'n': 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — 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
get_KNN_batch Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/ssaga_model.py
code served (permissive licence) · get_code("101d82f94bd38f5e")
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 int, shape (6, 3) 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)

# Parameters:
# - query_original_and_bert: tuple of 2 tensors, each [N1, d]
# - embedding_matrix_original_and_bert: tuple of 2 tensors, each [N2, d]
# - k: int (number of nearest neighbors)
# - device: torch device
# - batch_size: int

N1 = 6  # number of queries
N2 = 10  # number of candidates
d = 4   # embedding dimension
k = 3   # number of nearest neighbors
batch_size = 2

device = torch.device('cpu')

# Create query embeddings: tuple of 2 tensors [N1, d]
query_emb_0 = torch.randn(N1, d)
query_emb_1 = torch.randn(N1, d)
query_original_and_bert = (query_emb_0, query_emb_1)

# Create candidate embeddings: tuple of 2 tensors [N2, d]
candidate_emb_0 = torch.randn(N2, d)
candidate_emb_1 = torch.randn(N2, d)
embedding_matrix_original_and_bert = (candidate_emb_0, candidate_emb_1)

ARGS = {
    'query_original_and_bert': query_original_and_bert,
    'embedding_matrix_original_and_bert': embedding_matrix_original_and_bert,
    'k': k,
    'device': device,
    'batch_size': batch_size,
}
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 int, shape (6, 3) — shape (6, 3) 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_negative_samples_alignment Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/utils.py
code served (permissive licence) · get_code("f2461d0067c4446b")
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 1, dtype family int, shape (2,) 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)

ARGS = {
    "batch_size_each": 2,
    "num_entity": 10,
    "num_negative": None,
}
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 1, dtype family int, shape (2,) — shape (2,) 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_negative_samples_graph Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/utils.py
code served (permissive licence) · get_code("51910091f8bd88b1")
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 1, dtype family int, shape (2,) 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)

ARGS = {
    "batch_size_each": 2,
    "num_entity": 10,
}
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 1, dtype family int, shape (2,) — shape (2,) 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
hr2t_from_train_set Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/validate.py
code served (permissive licence) · get_code("acf0462dc6ad08ef")
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 pandas as pd
import tempfile
import os

# Create a temporary directory and TSV file for testing
temp_dir = tempfile.mkdtemp()
target_lang = "en"

# Create sample training data with head, relation, tail columns
train_data = {
    0: [1, 2, 3],  # head=1, relation=2, tail=3
    1: [1, 2, 4],  # head=1, relation=2, tail=4
    2: [5, 6, 7],  # head=5, relation=6, tail=7
    3: [1, 2, 3],  # duplicate triple
}

train_df = pd.DataFrame(train_data).T
train_file = os.path.join(temp_dir, f'{target_lang}-train.tsv')
train_df.to_csv(train_file, sep='\t', header=False, index=False)

ARGS = {
    'data_dir': temp_dir,
    'target_lang': target_lang,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — 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
k_hop_subgraph Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/graph_sampler.py
code served (permissive licence) · get_code("62d75dbae65c5fd5")
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 a simple graph with 5 nodes
# Edge index represents directed edges: (source, target)
edge_index = torch.tensor([
    [0, 1, 1, 2, 3],
    [1, 2, 3, 4, 4]
], dtype=torch.long)

# Define arguments for k_hop_subgraph
ARGS = {
    'node_idx': 0,
    'num_hops': 2,
    'edge_index': edge_index,
    'relabel_nodes': False,
    'num_nodes': 5,
    'flow': 'source_to_target'
}
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
l2distance Ran
ran on a synthesized input
amzn/ss-aga-kgc/src/ssaga_model.py
code served (permissive licence) · get_code("5b600a498c077665")
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 1, dtype family float, shape (2,) 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 two small tensors with distinct dimensions
# a: shape (2, 4) - batch of 2, dimension 4
# b: shape (2, 4) - batch of 2, dimension 4
a = torch.randn(2, 4)
b = torch.randn(2, 4)

ARGS = {
    "a": a,
    "b": b,
}
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 1, dtype family float, shape (2,) — shape (2,) 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
set_args Ran
ran on a synthesized input
amzn/ss-aga-kgc/run_model.py
code served (permissive licence) · get_code("9b545733db925243")
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 Namespace. 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 argparse

torch.manual_seed(0)

# Create a minimal args object with use_default set to True
args = argparse.Namespace()
args.use_default = True

# Add any other attributes that might be expected (even though set_args only checks use_default)
args.epoch10 = None
args.epoch11 = None
args.epoch2 = None
args.lr = None
args.align_lr = None
args.encoder_hdim_kg = None
args.encoder_hdim_align = None
args.dim = None
args.round = None
args.burin_in_epoch = None
args.generation_freq = None

ARGS = {"args": args}
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 Namespace — a Namespace output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Namespace 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
set_logger Ran
ran on a synthesized input
amzn/ss-aga-kgc/run_model.py
code served (permissive licence) · get_code("ac1cc817f2eb1c18")
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 int, ndim 0, dtype family int, 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 tempfile
import os
from types import SimpleNamespace

# Create a temporary directory for model_dir
model_dir = tempfile.mkdtemp()

# Create args object with required 'alias' attribute
args = SimpleNamespace(alias='test_experiment')

# Define ARGS dict with all parameters
ARGS = {
    'model_dir': model_dir,
    'args': args
}
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 int, ndim 0, dtype family int, 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
parse_args Not yet run
unverified (harvested, no recorded run)
amzn/ss-aga-kgc/run_model.py
code served (permissive licence) · get_code("8bf81439addc03d5")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 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 13 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

Predicting missing facts in a knowledge graph (KG) is crucial as modern KGs are far from complete. Due to labor-intensive human labeling, this phenomenon deteriorates when handling knowledge represented in various languages. In this paper, we explore multilingual KG completion, which leverages limited seed alignment as a bridge, to embrace the collective knowledge from multiple languages. However, language alignment used in prior works is still not fully exploited: (1) alignment pairs are treated equally to maximally push parallel entities to be close, which ignores KG capacity inconsistency; (2) seed alignment is scarce and new alignment identification is usually in a noisily unsupervised manner. To tackle these issues, we propose a novel self-supervised adaptive graph alignment (SS-AGA) method. Specifically, SS-AGA fuses all KGs as a whole graph by regarding alignment as a new edge type. As such, information propagation and noise influence across KGs can be adaptively controlled via relation-aware attention weights. Meanwhile, SS-AGA features a new pair generator that dynamically captures potential alignment pairs in a self-supervised paradigm. Extensive experiments on both the public multilingual DBPedia KG and newly-created industrial multilingual E-commerce KG empirically demonstrate the effectiveness of SS-AGA 1 .

For agents

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

get_harvested_code_for_paper("2203.14987")
get_code_for_paper("2203.14987")
have("2203.14987")

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

curl https://syntology.ai/api/ran/2203.14987.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.14987.svg)](https://syntology.ai/paper/2203.14987)

Connect an agent — have() is free.