SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1906.02505 · 2019

Fine-Grained Entity Typing in Hyperbolic Space

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 7 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/1906.02505.json · badge: /api/ran/1906.02505.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
nlpAThits/figet-hyperbolic-space canonical (classifier label)
classifier basis: census
7 ran · 1 no recorded run
FunctionStatusWhere it lives
cosine_distance Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/Predictor.py
code served (permissive licence) · get_code("0eca2f8d34319316")
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)

# The function cosine_distance takes two parameters: a and b
# Both are used with cos_sim_func which is torch.nn.CosineSimilarity(dim=0)
# CosineSimilarity expects two tensors and computes similarity along dim=0
# So a and b should be tensors with at least 1 dimension

# Create two small tensors with shape (4, 2) - batch size 2, dimension 4
a = torch.randn(4, 2)
b = torch.randn(4, 2)

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
f1 Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/evaluate.py
code served (permissive licence) · get_code("b9f6dc91f68b705d")
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 torch
import numpy as np

torch.manual_seed(0)

# Analyzing f1(p, r):
# - Takes two parameters: p and r
# - Performs arithmetic operations: comparison, multiplication, division
# - Both p and r are used in numerical operations
# - They should be numeric types (float or numeric tensors)

# Since the function uses these values in arithmetic and comparison,
# they can be floats, ints, or numeric tensors.
# Let's use floats as they're the simplest valid arguments.

p = 0.8
r = 0.6

ARGS = {
    "p": p,
    "r": r
}
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 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
hyperbolic_distance Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/hyperbolic.py
code served (permissive licence) · get_code("c863cbc3502f33f1")
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 torch
import numpy as np
import math

torch.manual_seed(0)

# The function takes three scalar or array arguments representing norms
# p_norm: norm of vector p
# q_norm: norm of vector q
# p_minus_q_norm: norm of (p - q)

# For hyperbolic distance, these should be values in [0, 1) range
# (they represent norms in hyperbolic space)

p_norm = 0.5
q_norm = 0.3
p_minus_q_norm = 0.4

ARGS = {
    "p_norm": p_norm,
    "q_norm": q_norm,
    "p_minus_q_norm": p_minus_q_norm,
}
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 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
hyperbolic_distance_torch Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/hyperbolic.py
code served (permissive licence) · get_code("d8c264bff46a42c1")
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. 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
import numpy as np

torch.manual_seed(0)

# Create small numpy arrays for p and q
# The function expects numpy arrays that will be converted to tensors
# Shape: (batch_size, embedding_dim) = (2, 4)
p = np.random.randn(2, 4).astype(np.float32)
q = np.random.randn(2, 4).astype(np.float32)

# Ensure values are in valid range for Poincare distance (within unit ball)
# Clamp to [-0.5, 0.5] to keep squared norms well below 1
p = np.clip(p, -0.5, 0.5)
q = np.clip(q, -0.5, 0.5)

ARGS = {
    'p': p,
    'q': q,
}
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 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
loose_macro Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/evaluate.py
code served (permissive licence) · get_code("3dfbdd021d8f9521")
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)

# The function loose_macro expects:
# true_and_prediction: an iterable of tuples (true_labels, predicted_labels)
# where each element is a tensor that supports:
#   - len() to get the count
#   - .item() on each element to get scalar values
#   - set operations and intersection

# Create sample data: list of (true_labels, predicted_labels) tuples
# Each label tensor contains indices that can be converted to items
true_and_prediction = [
    (torch.tensor([0, 1, 2]), torch.tensor([1, 2, 3])),
    (torch.tensor([0, 2]), torch.tensor([0, 1])),
    (torch.tensor([1, 3, 4]), torch.tensor([1, 4, 5, 6])),
]

ARGS = {
    "true_and_prediction": true_and_prediction,
}
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
sort_batch_by_length Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/model_utils.py
code served (permissive licence) · get_code("a1392b0f99913ab9")
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
from torch.autograd import Variable

torch.manual_seed(0)

# Create a small batch of sequences
# batch_size=2, sequence_length=4, feature_dim=3
batch_size = 2
seq_length = 4
feature_dim = 3

# Create tensor: batch_first format (batch_size, seq_length, feature_dim)
tensor_data = torch.randn(batch_size, seq_length, feature_dim)
tensor = Variable(tensor_data, requires_grad=False)

# Create sequence lengths: one length per batch element
# Use distinct values to make sorting observable
sequence_lengths_data = torch.LongTensor([4, 2])
sequence_lengths = Variable(sequence_lengths_data)

ARGS = {
    "tensor": tensor,
    "sequence_lengths": sequence_lengths,
}
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
strict Ran
ran on a synthesized input
nlpAThits/figet-hyperbolic-space/figet/evaluate.py
code served (permissive licence) · get_code("ac4451ffaa34240d")
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)

# The function expects true_and_prediction to be an iterable of tuples
# where each tuple contains (true_labels, predicted_labels)
# Both should be tensors with a .size() method and support == comparison

true_labels_1 = torch.tensor([0, 1, 2, 3])
predicted_labels_1 = torch.tensor([0, 1, 2, 3])

true_labels_2 = torch.tensor([1, 2, 3, 0])
predicted_labels_2 = torch.tensor([1, 2, 3, 0])

true_labels_3 = torch.tensor([0, 1, 0, 1])
predicted_labels_3 = torch.tensor([0, 0, 0, 1])

true_and_prediction = [
    (true_labels_1, predicted_labels_1),
    (true_labels_2, predicted_labels_2),
    (true_labels_3, predicted_labels_3),
]

ARGS = {
    "true_and_prediction": true_and_prediction
}
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
hyperbolic_distance_numpy Not yet run
unverified (harvested, no recorded run)
nlpAThits/figet-hyperbolic-space/figet/hyperbolic.py
code served (permissive licence) · get_code("92cb4e20f132bebc")
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, 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 8 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

How can we represent hierarchical information present in large type inventories for entity typing? We study the ability of hyperbolic embeddings to capture hierarchical relations between mentions in context and their target types in a shared vector space. We evaluate on two datasets and investigate two different techniques for creating a large hierarchical entity type inventory: from an expert-generated ontology and by automatically mining type co-occurrences. We find that the hyperbolic model yields improvements over its Euclidean counterpart in some, but not all cases. Our analysis suggests that the adequacy of this geometry depends on the granularity of the type inventory and the way hierarchical relations are inferred.

For agents

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

get_harvested_code_for_paper("1906.02505")
get_code_for_paper("1906.02505")
have("1906.02505")

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

curl https://syntology.ai/api/ran/1906.02505.json

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

[![Syntology run record](https://syntology.ai/api/ran/1906.02505.svg)](https://syntology.ai/paper/1906.02505)

Connect an agent — have() is free.