SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1906.07651 · 2019

Scheduled Sampling for Transformers

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 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; 1 honoured the contract we drafted, 1 violated it and 0 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/1906.07651.json · badge: /api/ran/1906.07651.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
deep-spin/scheduled-sampling-transformers canonical (classifier label)
classifier basis not recorded on the link
2 ran · 1 no recorded run
FunctionStatusWhere it lives
check_sru_requirement Ran
ran, violated the contract we drafted
deep-spin/scheduled-sampling-transformers/onmt/models/sru.py
code served (permissive licence) · get_code("0c93830510f645a3")
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 bool, ndim 0, dtype family float, shape () 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

torch.manual_seed(0)

ARGS = {
    'abort': False
}
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 bool, 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
count_features Ran
ran, honoured the contract we drafted
deep-spin/scheduled-sampling-transformers/preprocess.py
code served (permissive licence) · get_code("0c8cfd6442c243f1")
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 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 codecs
import tempfile
import os

# Create a temporary file with the expected format
# Format: whitespace-delimited tokens with │-delimited features within each token
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "corpus.txt")

# Write a sample corpus file
# First token has features separated by │
# Format: feature1│feature2│feature3 (which means 2 features, since we subtract 1)
with codecs.open(temp_file, 'w', 'utf-8') as f:
    f.write("word│pos│lemma other_token│tag\n")
    f.write("another│noun│base more_data\n")

ARGS = {
    "path": temp_file
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
none outside the standard library — 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
seq_linear Not yet run
unverified (harvested, no recorded run)
deep-spin/scheduled-sampling-transformers/onmt/modules/conv_multi_step_attention.py
pointer only (licence: MIT); the stored copy get_code serves, from Alex-Fabbri/Multi-News, is pointer only (licence: NOASSERTION) · get_code("0c3bddd7d3eaba02")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Alex-Fabbri/Multi-News, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 3 the same month, 3 later) and 0 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.

FunctionAlso present inEarliest-datedOthers named
check_sru_requirement 8 other papers
2 dated earlier, 1 the same month, 5 later
1709.02755 (2017-09)
Simple Recurrent Units for Highly Parallelizable Recurrence
1805.11462 (2018-05)
OpenNMT: Neural Machine Translation Toolkit
1906.06253 (2019-06)
A Simple and Effective Approach to Automatic Post-Editing with Transfer Learning
and 5 more not named here
seq_linear 8 other papers
1 dated earlier, 1 the same month, 6 later
1805.11462 (2018-05)
OpenNMT: Neural Machine Translation Toolkit
1906.06253 (2019-06)
A Simple and Effective Approach to Automatic Post-Editing with Transfer Learning
1910.02555 (2019-10)
Domain Differential Adaptation for Neural Machine Translation
and 5 more not named here
count_features 6 other papers
2 dated earlier, 1 the same month, 3 later
1805.11462 (2018-05)
OpenNMT: Neural Machine Translation Toolkit
1809.07358 (2018-09)
A Dataset for Document Grounded Conversations
1906.06253 (2019-06)
A Simple and Effective Approach to Automatic Post-Editing with Transfer Learning
and 3 more not named here

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

Scheduled sampling is a technique for avoiding one of the known problems in sequence-to-sequence generation: exposure bias. It consists of feeding the model a mix of the teacher forced embeddings and the model predictions from the previous step in training time. The technique has been used for improving the model performance with recurrent neural networks (RNN). In the Transformer model, unlike the RNN, the generation of a new word attends to the full sentence generated so far, not only to the last word, and it is not straightforward to apply the scheduled sampling technique. We propose some structural changes to allow scheduled sampling to be applied to Transformer architecture, via a two-pass decoding strategy. Experiments on two language pairs achieve performance close to a teacher-forcing baseline and show that this technique is promising for further exploration.

For agents

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

get_harvested_code_for_paper("1906.07651")
get_code_for_paper("1906.07651")
have("1906.07651")

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

curl https://syntology.ai/api/ran/1906.07651.json

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

[![Syntology run record](https://syntology.ai/api/ran/1906.07651.svg)](https://syntology.ai/paper/1906.07651)

Connect an agent — have() is free.