SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1611.01964 · 2016

Log-time and Log-space Extreme Classification

arXiv · PDF · Open in the Atlas

Code that ran

3 of 7 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 3 have no contract verdict; 0 failures of our instrument; 4 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/1611.01964.json · badge: /api/ran/1611.01964.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
kjasinska/ltls community (classifier label)
classifier basis not recorded on the link
3 ran · 4 no recorded run
FunctionStatusWhere it lives
compute_here_to_sink Ran
ran on a synthesized input
kjasinska/ltls/utils/graph.py
code served (permissive licence) · get_code("1eb74f74d08a375b")
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 torch
import numpy as np

torch.manual_seed(0)

# Analyzing the function:
# - adj: a dict-like object where adj[node] is iterable (contains successors)
# - n_classes: an integer that should equal hs[0] (the value at node 0)
# 
# The function builds a DAG where:
# - num_nodes = len(adj) + 1
# - sink = num_nodes - 1
# - Each node's hs value is the sum of its successors' hs values
# - hs[sink] = 1
# - The assertion checks that hs[0] == n_classes
#
# We need to construct a simple DAG that satisfies this.
# Let's create a small example:
# - adj has 3 nodes (indices 0, 1, 2)
# - num_nodes = 4, sink = 3
# - hs[3] = 1
# - hs[2] = hs[3] = 1 (node 2 points to sink)
# - hs[1] = hs[3] = 1 (node 1 points to sink)
# - hs[0] = hs[1] + hs[2] = 2 (node 0 points to 1 and 2)
# So n_classes should be 2

adj = {
    0: [1, 2],    # node 0 has successors 1 and 2
    1: [3],       # node 1 has successor 3 (the sink)
    2: [3],       # node 2 has successor 3 (the sink)
}

n_classes = 2

ARGS = {
    "adj": adj,
    "n_classes": n_classes,
}
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 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
load_dataset Ran
ran on a synthesized input
kjasinska/ltls/utils/io.py
code served (permissive licence) · get_code("043d5512de4d7487")
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 tempfile
import os
import numpy as np
from scipy.sparse import csr_matrix

# Create temporary SVMLight format files for testing
temp_dir = tempfile.mkdtemp()

def create_svmlight_file(filepath, n_samples=5, n_features=10, multilabel=False):
    """Create a simple SVMLight format file"""
    with open(filepath, 'w') as f:
        for i in range(n_samples):
            if multilabel:
                label = f"{i % 2},{(i + 1) % 2}"
            else:
                label = str(i % 3)
            features = " ".join([f"{j}:{np.random.rand():.4f}" for j in range(n_features)])
            f.write(f"{label} {features}\n")

# Create three sample files
path_train = os.path.join(temp_dir, "train.txt")
path_valid = os.path.join(temp_dir, "valid.txt")
path_test = os.path.join(temp_dir, "test.txt")

create_svmlight_file(path_train, n_samples=5, n_features=10, multilabel=False)
create_svmlight_file(path_valid, n_samples=3, n_features=10, multilabel=False)
create_svmlight_file(path_test, n_samples=2, n_features=10, multilabel=False)

# Define ARGS for the load_dataset function
ARGS = {
    'path_train': path_train,
    'n_features': 10,
    'path_valid': path_valid,
    'path_test': path_test,
    'multilabel': False,
    'classes_': None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, sklearn — 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
merge_rankings Ran
ran on a synthesized input
kjasinska/ltls/utils/ranking.py
code served (permissive licence) · get_code("6ddbbef5072d7751")
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
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# - temp_ranking: list of tuples where each tuple is (weight: float, path: list)
# - ranking_parent: list of tuples where each tuple is (weight: float, path: list)
# - ew_parent_current: float (edge weight)
# - parent: hashable value (used in tuple as (parent, current))
# - current: hashable value (used in tuple as (parent, current))
# - k: int (limit on new_ranking length)

# Build sample data
temp_ranking = [
    (1.5, [(0, 1)]),
    (2.0, [(0, 1), (1, 2)]),
    (3.5, [(0, 1), (1, 2), (2, 3)]),
]

ranking_parent = [
    (0.5, [(10, 11)]),
    (1.0, [(10, 11), (11, 12)]),
]

ew_parent_current = 0.75  # edge weight between parent and current

parent = 5
current = 6

k = 5

ARGS = {
    "temp_ranking": temp_ranking,
    "ranking_parent": ranking_parent,
    "ew_parent_current": ew_parent_current,
    "parent": parent,
    "current": current,
    "k": k,
}
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 — 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
count_edge_skips Not yet run
unverified (harvested, no recorded run)
kjasinska/ltls/utils/graph.py
code served (permissive licence) · get_code("11eb2b4c43fdcb99")
How it ran

no invocation recorded.

create_ranking1 Not yet run
unverified (harvested, no recorded run)
kjasinska/ltls/utils/ranking.py
code served (permissive licence) · get_code("dfd69c9dcb79e8ba")
How it ran

no invocation recorded.

create_ranking2 Not yet run
unverified (harvested, no recorded run)
kjasinska/ltls/utils/ranking.py
code served (permissive licence) · get_code("4013f47fb3f62ce4")
How it ran

no invocation recorded.

gengraph Not yet run
unverified (harvested, no recorded run)
kjasinska/ltls/utils/graph.py
code served (permissive licence) · get_code("ececaaa2fec797a3")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 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 7 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

We present LTLS, a technique for multiclass and multilabel prediction that can perform training and inference in logarithmic time and space. LTLS embeds large classification problems into simple structured prediction problems and relies on efficient dynamic programming algorithms for inference. We train LTLS with stochastic gradient descent on a number of multiclass and multilabel datasets and show that despite its small memory footprint it is often competitive with existing approaches.

For agents

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

get_harvested_code_for_paper("1611.01964")
get_code_for_paper("1611.01964")
have("1611.01964")

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

curl https://syntology.ai/api/ran/1611.01964.json

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

[![Syntology run record](https://syntology.ai/api/ran/1611.01964.svg)](https://syntology.ai/paper/1611.01964)

Connect an agent — have() is free.