SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.03079 · EMNLP · 2020

Blank Language Models

Regina Barzilay, Tommi Jaakkola, Tianxiao Shen, Victor Quach

arXiv · PDF · Open in the Atlas

Code that ran

7 of 8 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, 0 violated it and 6 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-21; 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/2002.03079.json · badge: /api/ran/2002.03079.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
Varal7/blank_language_model canonical (classifier label)
classifier basis: census
7 ran · 1 no recorded run
FunctionStatusWhere it lives
config_opt_schedule Ran
ran on a synthesized input
Varal7/blank_language_model/optimizer.py
code served (permissive licence) · get_code("efb8f3dfa32af36f")
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, 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 torch.nn as nn

torch.manual_seed(0)

# Create a simple model with parameters
model = nn.Linear(4, 3)
params = model.parameters()

# Create a minimal args object with all required attributes
class Args:
    def __init__(self):
        self.adam_betas = "(0.9, 0.999)"
        self.adam_eps = 1e-8
        self.weight_decay = 0.01
        self.lr = 0.001
        self.lr_schedule = 'triangular'
        self.warmup_steps = 100
        self.descend_steps = 100

args = Args()

ARGS = {
    'params': params,
    'args': args
}
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, ndim 2, dtype family float — a tuple 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
get_hparams Ran
ran on a synthesized input
Varal7/blank_language_model/utils.py
code served (permissive licence) · get_code("96be2538380bf7f7")
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 os
import tempfile
import yaml

# Create a temporary directory structure to mimic a checkpoint hierarchy
temp_dir = tempfile.mkdtemp()
model_dir = os.path.join(temp_dir, 'model_subdir')
os.makedirs(model_dir, exist_ok=True)

# Create hparams.yaml file at the parent level
hparams_data = {
    'learning_rate': 0.001,
    'batch_size': 32,
    'num_epochs': 10
}
hparams_path = os.path.join(temp_dir, 'hparams.yaml')
with open(hparams_path, 'w') as f:
    yaml.dump(hparams_data, f)

# Create a dummy checkpoint file in the model_subdir
checkpoint_path = os.path.join(model_dir, 'checkpoint.pt')
with open(checkpoint_path, 'w') as f:
    f.write('dummy checkpoint')

ARGS = {
    'checkpoint': checkpoint_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
yaml — 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
get_sinusoid_encoding_table Ran
ran, honoured the contract we drafted
Varal7/blank_language_model/transformer/Models.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from Matthewdowney18/Transformer_Dialogue, is pointer only (licence: NONE) · get_code("0f8f22937463d3c5")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from Matthewdowney18/Transformer_Dialogue, and GitHub detected no licence in that repository when it was checked. 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. A run is recorded.

repeat Ran
ran on a synthesized input
Varal7/blank_language_model/utils.py
code served (permissive licence) · get_code("fa5495dc0fc297ef")
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, 4) 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 simple function that will be applied repeatedly
def simple_function(x):
    """A simple function that applies a linear transformation"""
    return x * 2 + 1

# Define ARGS for the repeat function
ARGS = {
    "f": simple_function,
    "x": torch.tensor([[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]]),
    "n": 3
}
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, 4) — shape (2, 4) 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
select Ran
ran on a synthesized input
Varal7/blank_language_model/models/torch_utils.py
code served (permissive licence) · get_code("284f6884edd6b3fc")
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 1, dtype family int, shape (1,) 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

torch.manual_seed(0)

# Create a small logits tensor (batch_size=2, vocab_size=4)
logits = torch.randn(2, 4)

# Test both decode modes
ARGS = {
    'logits': logits,
    'decode': 'sample'
}
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 1, dtype family int, shape (1,) — shape (1,) 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_cross_entropy Ran
ran on a synthesized input
Varal7/blank_language_model/models/torch_utils.py
code served (permissive licence) · get_code("74dbb7aca393da78")
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.functional as F

torch.manual_seed(0)

# pred: should be a tensor with shape (batch, seq_len, vocab_size)
# When viewed as (-1, pred.size(-1)), it becomes (batch*seq_len, vocab_size)
batch_size = 2
seq_len = 3
vocab_size = 4

pred = torch.randn(batch_size, seq_len, vocab_size)

# gold: should be a tensor with shape (batch, seq_len) containing class indices
# Values should be in range [0, vocab_size) or equal to pad_idx
gold = torch.randint(0, vocab_size, (batch_size, seq_len))

# pad: padding index to ignore in the loss computation
pad = 0

ARGS = {
    "pred": pred,
    "gold": gold,
    "pad": pad
}
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
strip_eos Ran
ran on a synthesized input
Varal7/blank_language_model/utils.py
code served (permissive licence) · get_code("d0d0f038986aeb3e")
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

torch.manual_seed(0)

# The function strip_eos takes a list of sentences (lists or sequences)
# and removes everything from the '<eos>' token onwards (inclusive)
# It checks if '<eos>' is in each sent and uses .index() to find it

# Construct valid test data: a list of sentences where each sentence
# is a list of tokens (strings)
sents = [
    ['hello', 'world', '<eos>', 'extra', 'tokens'],
    ['foo', 'bar', 'baz'],
    ['a', '<eos>'],
    []
]

ARGS = {
    'sents': sents
}
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 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
new_arange Not yet run
unverified (harvested, no recorded run)
Varal7/blank_language_model/models/torch_utils.py
code served (permissive licence) · get_code("4c0673e3894b47a3")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 2 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 2 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.

FunctionAlso present inEarliest-datedOthers named
get_sinusoid_encoding_table 12 other papers
3 dated earlier, 8 later, 1 same year, order not determinable · 2 only as an assembly brick
1706.03762 (2017-06)
Attention Is All You Need
1810.04805 (2018-10)
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
1910.02481 (2019-10)
Learn to Explain Efficiently via Neural Logic Inductive Learning
and 9 more not named here
new_arange 6 other papers
6 later
2109.04084 (2021-09)
Thinking Clearly, Talking Fast: Concept-Guided Non-Autoregressive Generation for Open-Domain Dialogue Systems
2208.06073 (2022-08)
Conditional Antibody Design as 3D Equivariant Graph Translation
2302.01649 (2023-02)
Structure-informed Language Models Are Protein Designers
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 2026-11-21.

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 propose Blank Language Model (BLM), a model that generates sequences by dynamically creating and filling in blanks. The blanks control which part of the sequence to expand, making BLM ideal for a variety of text editing and rewriting tasks. The model can start from a single blank or partially completed text with blanks at specified locations. It iteratively determines which word to place in a blank and whether to insert new blanks, and stops generating when no blanks are left to fill. BLM can be efficiently trained using a lower bound of the marginal data likelihood. On the task of filling missing text snippets, BLM significantly outperforms all other baselines in terms of both accuracy and fluency. Experiments on style transfer and damaged ancient text restoration demonstrate the potential of this framework for a wide range of applications. 1

For agents

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

get_harvested_code_for_paper("2002.03079")
get_code_for_paper("2002.03079")
have("2002.03079")

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

curl https://syntology.ai/api/ran/2002.03079.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.03079.svg)](https://syntology.ai/paper/2002.03079)

Connect an agent — have() is free.