SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.00764 · EMNLP · 2021

SYSML: StYlometry with Structure and Multitask Learning: Implications for Darknet Forum Migrant Analysis

Srinivasan Parthasarathy, Pranav Maneriker, Yuntian He

arXiv · PDF · Open in the Atlas

Code that ran

5 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 1 failure of our instrument; 0 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/2104.00764.json · badge: /api/ran/2104.00764.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
pranavmaneriker/sysml marked official in Papers with Code
Papers with Code archive
5 ran · 0 no recorded run
FunctionStatusWhere it lives
ContextEmbeddingModel Ran
ran on a synthesized input · constructed an object; no result was computed
pranavmaneriker/sysml/src/custom_models.py
code served (permissive licence) · get_code("7254b45ac79447a0")
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, ContextEmbeddingModel: 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
from torch import nn

torch.manual_seed(0)

class MockContextTokenizer:
    def __init__(self, has_pretrained_emb=False):
        self.has_pretrained_emb = has_pretrained_emb
    
    def load_embeddings(self, emb_dim):
        # Return a small pretrained embedding tensor
        # Shape: (num_contexts, emb_dim)
        return torch.randn(5, emb_dim)

# Case 1: without pretrained embeddings
tokenizer_no_pretrained = MockContextTokenizer(has_pretrained_emb=False)

ARGS = {
    "num_contexts": 5,
    "context_tokenizer": tokenizer_no_pretrained,
    "emb_dim": 4
}
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 ContextEmbeddingModel — 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 ContextEmbeddingModel 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
CrossDatasetContextEmbedding Ran
ran on a synthesized input · constructed an object; no result was computed
pranavmaneriker/sysml/src/custom_models.py
code served (permissive licence) · get_code("11436cd4d6cb4d40")
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, CrossDatasetContextEmbedding: 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 numpy as np
from torch import nn

torch.manual_seed(0)

class MockContextTokenizer:
    def __init__(self, has_pretrained=False):
        self.has_pretrained_emb = has_pretrained
    
    def load_embeddings(self, emb_dim):
        return torch.randn(10, emb_dim)

class ContextEmbeddingModel(nn.Module):
    def __init__(self, num_contexts, context_tokenizer=None, emb_dim: int=32):
        super().__init__()
        if context_tokenizer and context_tokenizer.has_pretrained_emb:
            embedding = context_tokenizer.load_embeddings(emb_dim)
            self.embedding = nn.Embedding.from_pretrained(embedding)
        else:
            self.embedding = nn.Embedding(num_contexts, embedding_dim=emb_dim)
        self.final_dim = emb_dim

    def forward(self, d):
        return self.embedding(d)

# Create context embedding models with different sizes
tokenizer1 = MockContextTokenizer(has_pretrained=False)
model1 = ContextEmbeddingModel(num_contexts=5, context_tokenizer=tokenizer1, emb_dim=4)

tokenizer2 = MockContextTokenizer(has_pretrained=False)
model2 = ContextEmbeddingModel(num_contexts=7, context_tokenizer=tokenizer2, emb_dim=4)

tokenizer3 = MockContextTokenizer(has_pretrained=False)
model3 = ContextEmbeddingModel(num_contexts=6, context_tokenizer=tokenizer3, emb_dim=4)

context_embedding_models = [model1, model2, model3]

ARGS = {
    'context_embedding_models': context_embedding_models
}
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 CrossDatasetContextEmbedding — 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 CrossDatasetContextEmbedding 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
CrossEpisodeEmbModel Ran
ran on a synthesized input · constructed an object; no result was computed
pranavmaneriker/sysml/src/custom_models.py
code served (permissive licence) · get_code("b9f13a04fe03094b")
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, CrossEpisodeEmbModel: 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
import numpy as np

torch.manual_seed(0)

# Create a simple text model
class TextModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.embedding = nn.Embedding(100, 4)
        self.final_dim = 4
    
    def forward(self, x):
        return self.embedding(x).mean(dim=1)

# Create a simple time model
class TimeModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(1, 3)
        self.final_dim = 3
    
    def forward(self, x):
        return self.linear(x)

# Create a simple context tokenizer
class ContextTokenizer:
    def __init__(self):
        self.has_pretrained_emb = False

# Create ContextEmbeddingModel instances
context_tokenizer = ContextTokenizer()
emb_model_1 = nn.Module()
emb_model_1.embedding = nn.Embedding(5, 2)
emb_model_1.final_dim = 2

emb_model_2 = nn.Module()
emb_model_2.embedding = nn.Embedding(3, 2)
emb_model_2.final_dim = 2

context_embedding_models = [emb_model_1, emb_model_2]

# Create model instances
text_model = TextModel()
time_model = TimeModel()

# Create the batch dictionary
batch = {
    'texts': torch.randint(0, 100, (2, 3)),
    'times': torch.randn(2, 1),
    'context_offset': torch.randint(0, 8, (2,)),
    'lengths': torch.tensor([3, 2])
}

ARGS = {
    'text_model': text_model,
    'time_model': time_model,
    'context_embedding_models': context_embedding_models,
    'train_text': True,
    'train_time': True,
    'train_context': True
}
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 CrossEpisodeEmbModel — 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 CrossEpisodeEmbModel 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
either_train_or_eval Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
pranavmaneriker/sysml/src/custom_models.py
code served (permissive licence) · get_code("15f5035fb7caccff")
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
import torch.nn as nn

torch.manual_seed(0)

# Define a minimal model that can process the data
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create model instance
model = SimpleModel()

# Create small tensor data (batch_size=2, feature_dim=4)
data = torch.randn(2, 4)

# Create ARGS dict with all parameters
ARGS = {
    "model": model,
    "data": data,
    "train": True
}
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
get_final_dim Ran
ran, honoured the contract we drafted
pranavmaneriker/sysml/src/custom_models.py
code served (permissive licence) · get_code("04cc98da72d3b256")
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 torch
import torch.nn as nn

torch.manual_seed(0)

class ModelWithFinalDim(nn.Module):
    def __init__(self):
        super().__init__()
        self.final_dim = 64
        self.linear = nn.Linear(4, 64)
    
    def forward(self, x):
        return self.linear(x)

class ModelWithoutFinalDim(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 8)
    
    def forward(self, x):
        return self.linear(x)

model_with_final_dim = ModelWithFinalDim()
model_without_final_dim = ModelWithoutFinalDim()

ARGS = {
    'model': model_with_final_dim
}
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 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

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 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 5 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

Darknet market forums are frequently used to exchange illegal goods and services between parties who use encryption to conceal their identities. The Tor network is used to host these markets, which guarantees additional anonymization from IP and location tracking, making it challenging to link across malicious users using multiple accounts (sybils). Additionally, users migrate to new forums when one is closed further increasing the difficulty of linking users across multiple forums. We develop a novel stylometry-based multitask learning approach for natural language and model interactions using graph embeddings to construct low-dimensional representations of short episodes of user activity for authorship attribution. We provide a comprehensive evaluation of our methods across four different darknet forums demonstrating its efficacy over the state-of-the-art, with a lift of up to 2.5X on Mean Retrieval Rank and 2X on Recall@10.

For agents

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

get_harvested_code_for_paper("2104.00764")
get_code_for_paper("2104.00764")
have("2104.00764")

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

curl https://syntology.ai/api/ran/2104.00764.json

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

[![Syntology run record](https://syntology.ai/api/ran/2104.00764.svg)](https://syntology.ai/paper/2104.00764)

Connect an agent — have() is free.