SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.06979 · 2021

TSDAE: Using Transformer-based Sequential Denoising Auto-Encoder for Unsupervised Sentence Embedding Learning

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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 0 have no contract verdict; 3 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/2104.06979.json · badge: /api/ran/2104.06979.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
kwang2049/useb canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
ap_score Ran
ran; our contract draft was wrong, not the code
kwang2049/useb/useb/evaluators/askubuntu.py
code served (permissive licence) · get_code("98ef1f429528f0af")
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 numpy as np

# The function ap_score takes two lists:
# - relevant_list: a list of relevant item IDs
# - pred_list: a list of predicted item IDs (in ranked order)

# Both should be lists of hashable items (strings, ints, etc.)
# The function computes average precision metrics

# Create simple lists of item IDs
relevant_list = [1, 3, 5, 7]
pred_list = [1, 2, 3, 4, 5]

ARGS = {
    'relevant_list': relevant_list,
    'pred_list': pred_list
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — 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
rank_by_score Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
kwang2049/useb/useb/evaluators/askubuntu.py
code served (permissive licence) · get_code("f8c5243b757f193d")
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, 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.

import torch
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# - rank_by_score(candidates, scores)
# - It zips candidates and scores together
# - It sorts by kv[0][1], meaning candidates[i] must be indexable with [1]
# - It returns qid from (qid, _) tuples, so sorted_kvs contains (candidate, score) pairs
# - candidates is iterable and each element is indexable at position [1]
# - scores is iterable

# Create candidates: list of tuples where each tuple has at least 2 elements
# The function sorts by candidates[i][1], so we need that to be a numeric value
candidates = [
    ("query_1", 0.9),
    ("query_2", 0.7),
    ("query_3", 0.85),
]

# Create scores: list of numeric values (though they seem redundant given candidates structure)
scores = [0.5, 0.3, 0.8]

ARGS = {
    "candidates": candidates,
    "scores": scores,
}
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, 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
reciprocal_rank Ran
ran; our contract draft was wrong, not the code
kwang2049/useb/useb/evaluators/askubuntu.py
code served (permissive licence) · get_code("4ea966908d8fd441")
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:
# - relevant_list: a list/iterable that supports set operations and membership testing
# - pred_list: a list/iterable that supports enumeration and membership testing
# - The function finds the first item in pred_list that is in relevant_list
# - It needs at least one intersection between the two lists

# Create test data
relevant_list = [2, 4, 6, 8]
pred_list = [1, 3, 5, 2, 7]  # Contains 2 which is in relevant_list

ARGS = {
    'relevant_list': relevant_list,
    'pred_list': pred_list
}
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
run Not yet run
unverified (harvested, no recorded run)
kwang2049/useb/useb/evaluating.py
code served (permissive licence) · get_code("4b4d63cfd3f3c390")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 4 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

Learning sentence embeddings often requires a large amount of labeled data. However, for most tasks and domains, labeled data is seldom available and creating it is expensive. In this work, we present a new state-of-the-art unsupervised method based on pre-trained Transformers and Sequential Denoising Auto-Encoder (TSDAE) which outperforms previous approaches by up to 6.4 points. It can achieve up to 93.1% of the performance of in-domain supervised approaches. Further, we show that TSDAE is a strong domain adaptation and pre-training method for sentence embeddings, significantly outperforming other approaches like Masked Language Model. A crucial shortcoming of previous studies is the narrow evaluation: Most work mainly evaluates on the single task of Semantic Textual Similarity (STS), which does not require any domain knowledge. It is unclear if these proposed methods generalize to other domains and tasks. We fill this gap and evaluate TSDAE and other recent approaches on four different datasets from heterogeneous domains.

For agents

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

get_harvested_code_for_paper("2104.06979")
get_code_for_paper("2104.06979")
have("2104.06979")

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

curl https://syntology.ai/api/ran/2104.06979.json

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

[![Syntology run record](https://syntology.ai/api/ran/2104.06979.svg)](https://syntology.ai/paper/2104.06979)

Connect an agent — have() is free.