SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1705.08039 · 2017

Poincaré Embeddings for Learning Hierarchical Representations

arXiv · PDF · Open in the Atlas

Code that ran

6 of 14 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; 1 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 4 failures of our instrument; 8 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/1705.08039.json · badge: /api/ran/1705.08039.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
nishnik/poincare_embeddings reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 2 no recorded run
TatsuyaShirakawa/poincare-embedding reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
dalab/hyperbolic_cones extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
drewwilimitis/hyperbolic-learning extension (classifier label)
classifier basis: classifier_2025
0 ran · 6 no recorded run
FunctionStatusWhere it lives
dist Ran
ran, violated the contract we drafted
nishnik/poincare_embeddings/poincare.py
code served (permissive licence) · get_code("286b2d4f46b065fb")
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 float64, 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 numpy as np

# The function expects two numpy vectors (1D arrays)
# It computes a distance metric based on dot products
# The vectors should have the same dimension and represent points in a space
# where the norm is less than 1 (appears to be Poincaré disk model)

# Create two small vectors with norm < 1
vec1 = np.array([0.1, 0.2, 0.15])
vec2 = np.array([0.3, -0.1, 0.2])

ARGS = {
    'vec1': vec1,
    'vec2': vec2
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, 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
recover_tree_from_transitive_closure Ran
ran, honoured the contract we drafted
dalab/hyperbolic_cones/animation_train.py
code served (permissive licence) · get_code("b786803c03387b43")
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 list, ndim 2, 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.

# This function recovers direct parent-child relationships from a transitive closure
# of ancestor-descendant relationships in a tree.
# 
# The input is a list of tuples (ancestor, descendant) representing all ancestor-descendant
# pairs in a tree (including reflexive relations if present).
#
# Example: If we have a tree:
#     A
#    / \
#   B   C
#  /
# D
#
# The transitive closure would include:
# (A, A), (B, B), (C, C), (D, D) - reflexive
# (A, B), (A, C), (A, D) - A is ancestor of B, C, D
# (B, D) - B is ancestor of D
#
# The function should recover the direct edges: (A, B), (A, C), (B, D)

relations = [
    ('A', 'A'),  # reflexive
    ('B', 'B'),
    ('C', 'C'),
    ('D', 'D'),
    ('A', 'B'),  # A is ancestor of B (direct parent)
    ('A', 'C'),  # A is ancestor of C (direct parent)
    ('A', 'D'),  # A is ancestor of D (indirect, through B)
    ('B', 'D'),  # B is ancestor of D (direct parent)
]

ARGS = {
    'relations': relations
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 2, 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
transitive_closure Ran
ran; our contract draft was wrong, not the code
TatsuyaShirakawa/poincare-embedding/scripts/create_wordnet_noun_hierarchy.py
code served (permissive licence) · get_code("f4913cfd2afb2564")
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 set. 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.

class MockSynset:
    def __init__(self, name, pos_tag='n', hypernym_paths_list=None):
        self.name = name
        self.pos_tag = pos_tag
        self._hypernym_paths = hypernym_paths_list if hypernym_paths_list is not None else []
    
    def hypernym_paths(self):
        return self._hypernym_paths
    
    def pos(self):
        return self.pos_tag
    
    def __repr__(self):
        return f"MockSynset({self.name})"
    
    def __hash__(self):
        return hash(self.name)
    
    def __eq__(self, other):
        return isinstance(other, MockSynset) and self.name == other.name

# Create a hierarchy of synsets
root = MockSynset('entity', 'n')
physical_entity = MockSynset('physical_entity', 'n')
object_synset = MockSynset('object', 'n')
artifact = MockSynset('artifact', 'n')
vehicle = MockSynset('vehicle', 'n')
car = MockSynset('car', 'n')

# Set up hypernym paths for car
# Path 1: car -> vehicle -> artifact -> object -> physical_entity -> entity
path1 = [car, vehicle, artifact, object_synset, physical_entity, root]

# Path 2: car -> artifact -> object -> physical_entity -> entity (alternative path)
path2 = [car, artifact, object_synset, physical_entity, root]

car._hypernym_paths = [path1, path2]

# Set up hypernym paths for vehicle
vehicle._hypernym_paths = [[vehicle, artifact, object_synset, physical_entity, root]]

# Set up hypernym paths for artifact
artifact._hypernym_paths = [[artifact, object_synset, physical_entity, root]]

# Create synsets list
synsets = [car, vehicle]

ARGS = {
    'synsets': synsets
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 set — a set output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a set 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
transitive_isometry Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
TatsuyaShirakawa/poincare-embedding/scripts/plot_mammal_subtree.py
code served (permissive licence) · get_code("5d1f11cf5c981203")
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 function. 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.

# The function expects two tuples, each containing two numeric values (x, y coordinates)
# These represent points in the Poincaré disk model of hyperbolic geometry
# The function computes an isometry that maps t1 to t0

# t1 and t0 should be tuples of (x, y) where the point is in the unit disk
# For valid points in the Poincaré disk: x^2 + y^2 < 1

t1 = (0.3, 0.4)  # A point in the unit disk
t0 = (0.1, 0.2)  # Another point in the unit disk

ARGS = {
    't1': t1,
    't0': t0
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
update Ran
ran; our contract draft was wrong, not the code
nishnik/poincare_embeddings/poincare.py
code served (permissive licence) · get_code("b61dce6ea5a27ad7")
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 ndarray, ndim 1, dtype family float, shape (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 numpy as np
from math import sqrt

# The function expects:
# - emb: a numpy array (vector)
# - error_: a scalar value

# Create a small numpy array for emb
emb_param = np.array([0.1, 0.2, 0.3, 0.15])

# Create a scalar for error_
error_param = 0.5

ARGS = {
    'emb': emb_param,
    'error_': error_param
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (4,) — shape (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
update Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
nishnik/poincare_embeddings/poincare_adam.py
code served (permissive licence) · get_code("4a4bb7cf81fded2c")
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 ndarray, ndim 1, dtype family float, shape (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 numpy as np
from math import sqrt

# The function expects:
# - emb: a numpy array (vector)
# - error_: a scalar value

# Create a small numpy array for emb
emb_param = np.array([0.1, 0.2, 0.3, 0.4])

# Create a scalar error value
error_param = 0.5

ARGS = {
    'emb': emb_param,
    'error_': error_param
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (4,) — shape (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
exp_map Not yet run
unverified (harvested, no recorded run)
drewwilimitis/hyperbolic-learning/hyperbolic_kmeans/hkmeans.py
pointer only (licence: NONE) · get_code("4282578708af3132")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, drewwilimitis/hyperbolic-learning 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. No invocation recorded.

grad_fn Not yet run
unverified (harvested, no recorded run)
drewwilimitis/hyperbolic-learning/hyperbolic_svm/hsvm.py
pointer only (licence: NONE) · get_code("0d5d822d596da8c9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, drewwilimitis/hyperbolic-learning 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. No invocation recorded.

is_feasible Not yet run
unverified (harvested, no recorded run)
drewwilimitis/hyperbolic-learning/hyperbolic_svm/hsvm.py
pointer only (licence: NONE) · get_code("784529646199407c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, drewwilimitis/hyperbolic-learning 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. No invocation recorded.

loss_fn Not yet run
unverified (harvested, no recorded run)
drewwilimitis/hyperbolic-learning/hyperbolic_svm/hsvm.py
pointer only (licence: NONE) · get_code("d580486bed607544")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, drewwilimitis/hyperbolic-learning 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. No invocation recorded.

minkowski_distance_gradient Not yet run
unverified (harvested, no recorded run)
drewwilimitis/hyperbolic-learning/hyperbolic_kmeans/hkmeans.py
pointer only (licence: NONE) · get_code("3e8d29824b777e6a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, drewwilimitis/hyperbolic-learning 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. No invocation recorded.

minkowski_loss_gradient Not yet run
unverified (harvested, no recorded run)
drewwilimitis/hyperbolic-learning/hyperbolic_kmeans/hkmeans.py
pointer only (licence: NONE) · get_code("f97c938afe95df01")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, drewwilimitis/hyperbolic-learning 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. No invocation recorded.

partial_der Not yet run
unverified (harvested, no recorded run)
nishnik/poincare_embeddings/poincare.py
code served (permissive licence) · get_code("aeb157936be2be1e")
How it ran

no invocation recorded.

partial_der Not yet run
unverified (harvested, no recorded run)
nishnik/poincare_embeddings/poincare_adam.py
code served (permissive licence) · get_code("100812a0370f139e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 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

Representation learning has become an invaluable approach for learning from symbolic data such as text and graphs. However, while complex symbolic datasets often exhibit a latent hierarchical structure, state-of-the-art methods typically learn embeddings in Euclidean vector spaces, which do not account for this property. For this purpose, we introduce a new approach for learning hierarchical representations of symbolic data by embedding them into hyperbolic space -- or more precisely into an n-dimensional Poincaré ball. Due to the underlying hyperbolic geometry, this allows us to learn parsimonious representations of symbolic data by simultaneously capturing hierarchy and similarity. We introduce an efficient algorithm to learn the embeddings based on Riemannian optimization and show experimentally that Poincaré embeddings outperform Euclidean embeddings significantly on data with latent hierarchies, both in terms of representation capacity and in terms of generalization ability.

For agents

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

get_harvested_code_for_paper("1705.08039")
get_code_for_paper("1705.08039")
have("1705.08039")

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

curl https://syntology.ai/api/ran/1705.08039.json

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

[![Syntology run record](https://syntology.ai/api/ran/1705.08039.svg)](https://syntology.ai/paper/1705.08039)

Connect an agent — have() is free.