SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.03808 · EMNLP · 2021

Smelting Gold and Silver for Improved Multilingual AMR-to-Text Generation

Iryna Gurevych, Yue Zhang, Jonas Pfeiffer, Leonardo Ribeiro

arXiv · PDF · Open in the Atlas

Code that ran

11 of 11 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 5 have no contract verdict; 5 failures 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/2109.03808.json · badge: /api/ran/2109.03808.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
UKPLab/m-AMR2Text canonical (classifier label)
classifier basis: census
11 ran · 0 no recorded run
FunctionStatusWhere it lives
booleans_processing Ran
ran on a synthesized input
UKPLab/m-AMR2Text/transformers/modeling_tf_utils.py
code served (permissive licence) · get_code("3c4b7e11935f3cf8")
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 tensorflow as tf

class MockConfig:
    """Stand-in PretrainedConfig with required boolean attributes"""
    def __init__(self):
        self.output_attentions = False
        self.output_hidden_states = False
        self.return_dict = True
        self.use_cache = False

config = MockConfig()

kwargs = {
    'output_attentions': None,
    'output_hidden_states': None,
    'return_dict': None,
    'use_cache': None,
}

ARGS = {
    'config': config,
}

ARGS.update(kwargs)
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — 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
convert_tf_weight_name_to_pt_weight_name Ran
ran on a synthesized input
UKPLab/m-AMR2Text/transformers/modeling_tf_pytorch_utils.py
code served (permissive licence) · get_code("5f5d99a0dee8c2b1")
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 1, 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 re

def convert_tf_weight_name_to_pt_weight_name(tf_name, start_prefix_to_remove=''):
    """
    Convert a TF 2.0 model variable name in a pytorch model weight name.

    Conventions for TF2.0 scopes -> PyTorch attribute names conversions:

        - '$1___$2' is replaced by $2 (can be used to duplicate or remove layers in TF2.0 vs PyTorch)
        - '_._' is replaced by a new level separation (can be used to convert TF2.0 lists in PyTorch nn.ModulesList)

    return tuple with:

        - pytorch model weight name
        - transpose: boolean indicating whether TF2.0 and PyTorch weights matrices are transposed with regards to each
          other
    """
    tf_name = tf_name.replace(':0', '')
    tf_name = re.sub('/[^/]*___([^/]*)/', '/\\1/', tf_name)
    tf_name = tf_name.replace('_._', '/')
    tf_name = re.sub('//+', '/', tf_name)
    tf_name = tf_name.split('/')
    if len(tf_name) > 1:
        tf_name = tf_name[1:]
    transpose = bool(tf_name[-1] in ['kernel', 'pointwise_kernel', 'depthwise_kernel'] or 'emb_projs' in tf_name or 'out_projs' in tf_name)
    if tf_name[-1] == 'kernel' or tf_name[-1] == 'embeddings' or tf_name[-1] == 'gamma':
        tf_name[-1] = 'weight'
    if tf_name[-1] == 'beta':
        tf_name[-1] = 'bias'
    if tf_name[-1] == 'pointwise_kernel' or tf_name[-1] == 'depthwise_kernel':
        tf_name[-1] = tf_name[-1].replace('_kernel', '.weight')
    tf_name = '.'.join(tf_name)
    if start_prefix_to_remove:
        tf_name = tf_name.replace(start_prefix_to_remove, '', 1)
    return (tf_name, transpose)


ARGS = {
    'tf_name': 'layer/dense/kernel:0',
    'start_prefix_to_remove': ''
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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 tuple, ndim 1, 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
find_pruneable_heads_and_indices Ran
ran; our contract draft was wrong, not the code
UKPLab/m-AMR2Text/transformers/modeling_utils.py
code served (permissive licence) · get_code("6d254a1b39c650ff")
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)

# Small example values for the function parameters
heads_to_prune = [1, 3]  # List of head indices to prune
n_heads = 8  # Total number of heads in the model
head_size = 4  # Size of each head
already_pruned = {0, 2}  # Set of heads already pruned

ARGS = {
    "heads": heads_to_prune,
    "n_heads": n_heads,
    "head_size": head_size,
    "already_pruned_heads": already_pruned,
}
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
gelu_fast Ran
ran; our contract draft was wrong, not the code
UKPLab/m-AMR2Text/transformers/activations.py
code served (permissive licence) · get_code("a4475703ff58ecf9")
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 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

torch.manual_seed(0)

# gelu_fast takes a single parameter 'x' which is used in arithmetic operations
# It should be a tensor that supports multiplication, addition, and tanh operations
x = torch.randn(2, 4)

ARGS = {
    "x": x
}
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 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
gelu_fast Ran
ran on a synthesized input
UKPLab/m-AMR2Text/transformers/activations_tf.py
code served (permissive licence) · get_code("ac1bc5cb23b3f2ec")
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 EagerTensor, 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 tensorflow as tf
import numpy as np

# Create a small input tensor for testing
# Using shape (2, 4) - batch size 2, feature dim 4
x_input = tf.constant([
    [0.1, 0.2, 0.3, 0.4],
    [-0.5, 1.0, -0.1, 0.8]
], dtype=tf.float32)

ARGS = {
    "x": x_input
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, 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
gelu_new Ran
ran; our contract draft was wrong, not the code
UKPLab/m-AMR2Text/transformers/activations.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from astro-jon/LCWL, is pointer only (licence: NONE) · get_code("45bb87451230d5e8")
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 astro-jon/LCWL, 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.

generate_edge_tensors Ran
ran on a synthesized input
UKPLab/m-AMR2Text/utils_graphadapt.py
code served (permissive licence) · get_code("c0726694b5cef1c0")
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
import numpy as np

torch.manual_seed(0)

# Analyzing the function:
# - graphs: iterable of strings, each string contains space-separated edge tuples
# - max_source_length: integer used for filtering edges
# - Each edge tuple has format: "(e2,e1,label)" where e1, e2 are integers and label is 'd' or 'r'

# Create sample graph strings with edge tuples
graph1 = "(2,0,d) (3,1,r) (1,2,d)"
graph2 = "(1,0,r) (2,1,d)"

graphs = [graph1, graph2]
max_source_length = 5

ARGS = {
    "graphs": graphs,
    "max_source_length": max_source_length
}
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 — 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
generate_token_graph Ran
ran on a synthesized input
UKPLab/m-AMR2Text/utils_graphadapt.py
code served (permissive licence) · get_code("f128d1c1edf9da40")
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 1, 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

class SimpleTokenizer:
    """Mock tokenizer that simulates sentencepiece tokenization."""
    def __init__(self):
        self.special_tokens = {':quant-of', ':snt4', ':op3', ':century', ':prep-by'}
        
    def tokenize(self, text):
        """Tokenize text with word piece logic using ▁ prefix for word starts."""
        tokens = []
        words = text.split()
        for word in words:
            if word in new_tokens:
                # Special AMR tokens stay as-is
                tokens.append(word)
            else:
                # Regular words: first token gets ▁ prefix, rest are subwords
                sub_tokens = list(word)
                if sub_tokens:
                    tokens.append('▁' + sub_tokens[0])
                    for char in sub_tokens[1:]:
                        tokens.append(char)
        return tokens

# Create mock data
new_tokens = {':quant-of', ':snt4', ':op3', ':century', ':prep-by', ':prep-amid', ':extent-of', ':prep-as', ':ord-of', ':prep-into', ':destination-of', ':year', ':prep-on-of', ':prep-for-of', ':part-of', ':prep-in-addition-to', ':medium-of', ':prep-to-of', ':frequency-of', ':prep-in', ':subset-of', ':instrument', ':op15', ':ARG6', ':ARG3', ':prep-on-behalf', ':prep-with', ':polite', ':ARG2', ':dayperiod-of', ':ARG0-of', ':value-of', ':op11', ':year2', ':prep-among', ':prep-against', ':prep-without-of', ':op18', ':quarter', ':manner', ':op14', ':snt10', ':ARG2-of', ':op1-of', ':topic', ':beneficiary-of', ':op16', ':snt5', ':prep-out-of', ':ARG5', ':purpose', ':frequency', ':purpose-of', ':condition-of', ':polite-of', ':subevent', ':condition', ':ARG7', ':prep-from', ':unit-of', ':month', ':name', ':source-of', ':age-of', ':op8', ':prep-at', ':poss', ':snt11', ':direction-of', ':path', ':compared-to', ':example', ':prep-under-of', ':op5-of', ':wiki', ':snt9', ':timezone', ':extent', ':example-of', ':value', ':<UNK>-of', ':concession', ':op2-of', ':snt2', ':part', ':op13', ':op17', ':ARG5-of', ':calendar', ':consist-of', ':li', ':prep-on-behalf-of', ':duration', ':beneficiary', ':polarity-of', ':<UNK>', ':source', ':op19', ':day', ':ARG8', ':s', ':domain-of', ':degree-of', ':op9', ':snt3', ':P', ':op5', ':prep-on', ':prep-along-with', ':op7', ':ARG6-of', ':name-of', ':op12', ':accompanier-of', ':medium', ':op20', ':snt1', ':op1', ':domain', ':op4', ':consist', ':instrument-of', ':snt8', ':prep-as-of', ':ord', ':op2', ':dayperiod', ':prep-for', ':mod', ':location', ':op10', ':ARG1', ':topic-of', ':snt6', ':season', ':quant', ':ARG4-of', ':age', ':subevent-of', ':time-of', ':polarity', ':location-of', ':conj-as-if', ':range', ':prep-to', ':op3-of', ':destination', ':duration-of', ':time', ':mode', ':mod-of', ':ARG0', ':manner-of', ':ARG9', ':range-of', ':direction', ':path-of', ':ARG1-of', ':decade', ':scale', ':degree', ':prep-under', ':li-of', ':prep-in-of', ':prep-with-of', ':op6', ':unit', ':ARG4', ':op4-of', ':snt7', ':prep-toward', ':prep-against-of', ':prep-without', ':year-of', ':accompanier', ':poss-of', ':weekday', ':era', ':concession-of', ':ARG3-of'}

# Create simple inputs - AMR-like text with special tokens
inputs = [
    ":op1 want :ARG0 boy :ARG1 go",
    ":op2 like :ARG0 girl :op1 dance"
]

# Create corresponding graphs with edges
# Format: (source_node_idx,target_node_idx,relation_label)
graphs = [
    "(0,1,:instance) (1,2,:ARG0) (1,3,:ARG1)",
    "(0,1,:instance) (1,2,:ARG0) (1,3,:op1)"
]

# Create tokenizer
tokenizer = SimpleTokenizer()

ARGS = {
    "inputs": inputs,
    "graphs": graphs,
    "tokenizer": tokenizer
}
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, ndim 1, 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
keras_serializable Ran
ran; our contract draft was wrong, not the code
UKPLab/m-AMR2Text/transformers/modeling_tf_utils.py
code served (permissive licence) · get_code("756e9ec5f7912d3e")
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 type. 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 tensorflow as tf
import functools

class PretrainedConfig:
    """Stand-in for PretrainedConfig"""
    def __init__(self, **kwargs):
        self.params = kwargs
    
    def to_dict(self):
        return self.params.copy()
    
    @classmethod
    def from_dict(cls, config_dict):
        return cls(**config_dict)

class TestKerasLayer(tf.keras.layers.Layer):
    """Minimal Keras Layer subclass for testing"""
    config_class = PretrainedConfig
    
    def __init__(self, config=None, **kwargs):
        super().__init__(**kwargs)
        self._config = config
        self._kwargs = kwargs
    
    def get_config(self):
        return super().get_config()
    
    def call(self, inputs):
        return inputs

# Mark get_config as default so the decorator will override it
TestKerasLayer.get_config._is_default = True

ARGS = {
    'cls': TestKerasLayer
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
mish Ran
ran, honoured the contract we drafted
UKPLab/m-AMR2Text/transformers/activations_tf.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from VITA-Group/READ-ME, is pointer only (licence: NONE) · get_code("cc8c8c3ebf0c343f")
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 VITA-Group/READ-ME, 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.

mish Ran
ran; our contract draft was wrong, not the code
UKPLab/m-AMR2Text/transformers/activations.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from thu-coai/Emotional-Support-Conversation, is pointer only (licence: NOASSERTION) · get_code("f889d53c68539013")
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 thu-coai/Emotional-Support-Conversation, 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. A run is recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 8 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 0 the same month, 7 later) and 3 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
mish 25 other papers
4 dated earlier, 21 later
2007.07779 (2020-07)
AdapterHub: A Framework for Adapting Transformers
2104.08400 (2021-04)
Structure-Aware Abstractive Conversation Summarization via Discourse and Action Graphs
2106.00420 (2021-06)
Dialogue-oriented Pre-training
and 22 more not named here
gelu_new 24 other papers
6 dated earlier, 18 later
2004.10964 (2020-04)
Don't Stop Pretraining: Adapt Language Models to Domains and Tasks
2104.08066 (2021-04)
Effect of Visual Extensions on Natural Language Understanding in Vision-and-Language Models
2104.08400 (2021-04)
Structure-Aware Abstractive Conversation Summarization via Discourse and Action Graphs
and 21 more not named here
gelu_fast 22 other papers
6 dated earlier, 16 later
2010.03957 (2020-10)
Transformers for Modeling Physical Systems
2104.08066 (2021-04)
Effect of Visual Extensions on Natural Language Understanding in Vision-and-Language Models
2104.08400 (2021-04)
Structure-Aware Abstractive Conversation Summarization via Discourse and Action Graphs
and 19 more not named here
find_pruneable_heads_and_indices 11 other papers
3 dated earlier, 7 later, 1 same year, order not determinable
2004.10964 (2020-04)
Don't Stop Pretraining: Adapt Language Models to Domains and Tasks
2021.emnlp-main.154 (2021 (year only))
Reconstruction Attack on Instance Encoding for Language Understanding
2104.08066 (2021-04)
Effect of Visual Extensions on Natural Language Understanding in Vision-and-Language Models
and 8 more not named here
keras_serializable 8 other papers
8 later
2303.14582 (2023-03)
Identification of Negative Transfers in Multitask Learning Using Surrogate Models
2310.05107 (2023-10)
OV-PARTS: Towards Open-Vocabulary Part Segmentation
2310.09832 (2023-10)
Merging Experts into One: Improving Computational Efficiency of Mixture of Experts
and 5 more not named here
mish 4 other papers
2 dated earlier, 2 later
2104.08066 (2021-04)
Effect of Visual Extensions on Natural Language Understanding in Vision-and-Language Models
2107.00910 (2021-07)
Learned Token Pruning for Transformers
2405.15179 (2024-05)
VB-LoRA: Extreme Parameter Efficient Fine-Tuning with Vector Banks
and 1 more not named here
convert_tf_weight_name_to_pt_weight_name 3 other papers
3 later
2022.findings-emnlp.427 (2022 (year only))
2203.10652 (2022-03)
Continual Sequence Generation with Adaptive Compositional Modules
2303.14582 (2023-03)
Identification of Negative Transfers in Multitask Learning Using Surrogate Models
gelu_fast 1 other paper
1 dated earlier
2107.00910 (2021-07)
Learned Token Pruning for Transformers

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

Recent work on multilingual AMR-to-text generation has exclusively focused on data augmentation strategies that utilize silver AMR. However, this assumes a high quality of generated AMRs, potentially limiting the transferability to the target task. In this paper, we investigate different techniques for automatically generating AMR annotations, where we aim to study which source of information yields better multilingual results. Our models trained on gold AMR with silver (machine translated) sentences outperform approaches which leverage generated silver AMR. We find that combining both complementary sources of information further improves multilingual AMR-to-text generation. Our models surpass the previous state of the art for German, Italian, Spanish, and Chinese by a large margin. 1

For agents

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

get_harvested_code_for_paper("2109.03808")
get_code_for_paper("2109.03808")
have("2109.03808")

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

curl https://syntology.ai/api/ran/2109.03808.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.03808.svg)](https://syntology.ai/paper/2109.03808)

Connect an agent — have() is free.