SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.14449 · NeurIPS · 2021

One Loss for All: Deep Hashing with a Single Cosine Similarity based Learning Objective

Tao Xiang, Jiun Hoe, Kam Woh Ng, Tianyu Zhang, Chee Chan, Yi-Zhe Song

arXiv · PDF · Open in the Atlas

Code that ran

4 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 2 failures of our instrument; 2 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/2109.14449.json · badge: /api/ran/2109.14449.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
kamwoh/orthohash canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
cisiplab/cisip-fire community
Papers with Code archive
2 ran · 1 no recorded run
FunctionStatusWhere it lives
OrthoCosLoss Ran
ran on a synthesized input · constructed an object; no result was computed
cisiplab/cisip-fire/functions/loss/orthocos.py
code served (permissive licence) · get_code("51f777eea1887485")
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 an instance of its own class, OrthoCosLoss: the run constructed the object and did not call it, so no result was computed. 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 torch.nn as nn

torch.manual_seed(0)

# Setup for OrthoCosLoss - define all constructor parameters
ARGS = {
    'ce': 1.0,
    'multiclass': False,
    's': 8,
    'm': 0.2,
    'quan': 0.5,
    'quan_type': 'cs',
    'multiclass_loss': 'label_smoothing',
}
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 OrthoCosLoss — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a OrthoCosLoss 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
OrthoHashLoss Ran
ran on a synthesized input
kamwoh/orthohash/functions/loss/orthohash.py
code served (permissive licence) · get_code("2ea411f078599d51")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 0, 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
import torch.nn as nn
import torch.nn.functional as F

# INIT_ARGS for OrthoHashLoss.__init__
INIT_ARGS = {
    'ce': 1.0,
    's': 8,
    'm': 0.2,
    'm_type': 'cos',
    'multiclass': True,
    'quan': 0.5,
    'quan_type': 'cs',
    'multiclass_loss': 'label_smoothing'
}

# ARGS for OrthoHashLoss.forward
batch_size = 2
num_classes = 5
code_dim = 8

# Create logits (normalized for cosine similarity interpretation)
logits = torch.randn(batch_size, num_classes)
logits = F.normalize(logits, p=2, dim=1)  # Normalized to simulate cosine similarities

# Create code_logits (hash codes before binarization)
code_logits = torch.randn(batch_size, code_dim)

# Create multi-hot labels (multiclass=True)
labels = torch.zeros(batch_size, num_classes)
labels[0, [0, 2, 3]] = 1.0  # First sample has classes 0, 2, 3
labels[1, [1, 4]] = 1.0      # Second sample has classes 1, 4

ARGS = {
    'logits': logits,
    'code_logits': code_logits,
    'labels': labels,
    'onehot': True
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 0, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
get_imbalance_mask Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
cisiplab/cisip-fire/functions/loss/orthocos.py
code served (permissive licence) · get_code("13450b62dcee2e2a")
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

torch.manual_seed(0)

# Create small tensors for sigmoid_logits and labels
batch_size = 2
num_classes = 4

# sigmoid_logits: should be logits that can be compared with threshold
sigmoid_logits = torch.rand(batch_size, num_classes)

# labels: binary labels (0 or 1) for multi-label classification
labels = torch.randint(0, 2, (batch_size, num_classes))

# nclass: number of classes
nclass = num_classes

# threshold: for binary classification
threshold = 0.7

# imbalance_scale: scaling factor for imbalanced classes
imbalance_scale = -1

ARGS = {
    "sigmoid_logits": sigmoid_logits,
    "labels": labels,
    "nclass": nclass,
    "threshold": threshold,
    "imbalance_scale": imbalance_scale,
}
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
one_hot Ran
ran; our contract draft was wrong, not the code
kamwoh/orthohash/utils/datasets.py
code served (permissive licence) · get_code("d2b1a5f7c60e27b5")
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 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.

import torch

torch.manual_seed(0)

# The function one_hot takes a single parameter: nclass
# nclass is used as the second argument to torch.nn.functional.one_hot
# It specifies the number of classes for the one-hot encoding

ARGS = {
    "nclass": 5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — 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
BaseClassificationLoss Not yet run
unverified (harvested, no recorded run)
cisiplab/cisip-fire/functions/loss/orthocos.py
code served (permissive licence) · get_code("bedb4a705f71cf90")
How it ran

no invocation recorded.

BaseClassificationLoss Not yet run
unverified (harvested, no recorded run)
kamwoh/orthohash/functions/loss/orthohash.py
code served (permissive licence) · get_code("2fe62c1d3f1d5224")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 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 6 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

A deep hashing model typically has two main learning objectives: to make the learned binary hash codes discriminative and to minimize a quantization error. With further constraints such as bit balance and code orthogonality, it is not uncommon for existing models to employ a large number (>4) of losses. This leads to difficulties in model training and subsequently impedes their effectiveness. In this work, we propose a novel deep hashing model with only a single learning objective. Specifically, we show that maximizing the cosine similarity between the continuous codes and their corresponding binary orthogonal codes can ensure both hash code discriminativeness and quantization error minimization. Further, with this learning objective, code balancing can be achieved by simply using a Batch Normalization (BN) layer and multi-label classification is also straightforward with label smoothing. The result is an one-loss deep hashing model that removes all the hassles of tuning the weights of various losses. Importantly, extensive experiments show that our model is highly effective, outperforming the state-of-the-art multi-loss hashing models on three large-scale instance retrieval benchmarks, often by significant margins. Code is available at https://github.com/kamwoh/orthohash

For agents

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

get_harvested_code_for_paper("2109.14449")
get_code_for_paper("2109.14449")
have("2109.14449")

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

curl https://syntology.ai/api/ran/2109.14449.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.14449.svg)](https://syntology.ai/paper/2109.14449)

Connect an agent — have() is free.