SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.08253 · 2021

A Syntax-Guided Edit Decoder for Neural Program Repair

arXiv · PDF · Open in the Atlas

Code that ran

12 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 8 have no contract verdict; 3 failures of our instrument; 2 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/2106.08253.json · badge: /api/ran/2106.08253.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
zysszy/TreeGen canonical (classifier label)
classifier basis: census
12 ran · 2 no recorded run
FunctionStatusWhere it lives
J_findtheson_name Ran
ran; our contract draft was wrong, not the code
zysszy/TreeGen/predict.py
code served (permissive licence) · get_code("a2a49f015c3f5c67")
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 numpy as np

# Analyze the function:
# - l: a list that is indexed with integers (len(l) is called, l[nowsite] is compared to '^')
# - site: an integer used as starting index
# The function traverses a list looking for characters, counting '^' to decrement and other chars to increment
# It collects characters when count == 1
# This appears to be parsing some kind of tree structure where '^' marks the end of a node

# Create a simple list with some characters and '^' markers
# Example: a simple tree-like structure
l = ['a', 'b', '^', 'c', 'd', '^', '^']

# Starting site at index 0
site = 0

ARGS = {
    'l': l,
    'site': site
}
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 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
J_isend Ran
ran, violated the contract we drafted
zysszy/TreeGen/predict.py
code served (permissive licence) · get_code("8de998fbb776b7de")
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 bool, ndim 0, dtype family float, 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.

# Analyze the function J_isend to understand what it needs:
# - Parameter 'l': a list that is indexed with integers, contains strings like '^', 'End', and node names
# - Parameter 'site': an integer index into list l
# - Uses global lists: J_HasSon and J_NeedsEnd

# Set up the global lists
J_HasSon = ['NodeA', 'NodeB', 'NodeC']
J_NeedsEnd = ['NodeB', 'NodeC']

# Create a test list l that represents a tree structure
# The function appears to parse a tree structure where '^' marks end of children
# Example structure: ['NodeA', 'NodeB', 'Child1', '^', 'End', '^']
l = ['NodeA', 'NodeB', 'Child1', '^', 'End', '^', 'NodeC', 'Child2', '^']

# site is an index into the list
site = 0

# Construct ARGS
ARGS = {
    'l': l,
    'site': site
}
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 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
J_readlist Ran
ran; our contract draft was wrong, not the code
zysszy/TreeGen/predict.py
code served (permissive licence) · get_code("b72343ca8c77e92f")
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 io

in_file = io.StringIO("node1 node2 node3 node_gen ^ node4\n42\n")

ARGS = {
    "in_file": in_file
}
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 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
ast2code Ran
ran on a synthesized input
zysszy/TreeGen/HS-B/eval/ast2code.py
code served (permissive licence) · get_code("24b755519861ca11")
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 str. 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

# Analyze the function to understand what it needs
# ast2code takes 3 parameters: ast, d, fa
# - ast: appears to be a list where ast[0] is a string (node type) and ast[1:] contains child nodes
# - d: appears to be an integer (depth level for indentation)
# - fa: appears to be a string (parent/family node type)

# The function recursively processes AST nodes and generates code strings
# We need to create a minimal valid AST structure

# Create a simple AST node structure
# An AST node is a list where:
# - ast[0] is the node type (string)
# - ast[1:] contains child nodes, marked with '^' for closing

# Example: a simple assignment node
ast_node = [
    'Assign',           # node type
    'Name',             # child 1
    'x',                # child 1 content
    '^',                # closing marker for child 1
    'value',            # child 2
    'Constant',         # child 2 content
    '42',               # child 2 content
    '^'                 # closing marker for child 2
]

# Parameters:
# - ast: the AST node structure
# - d: depth (integer, used for indentation)
# - fa: family/parent node type (string)

ARGS = {
    'ast': ast_node,
    'd': 0,
    'fa': 'Module'
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
find_father Ran
ran on a synthesized input
zysszy/TreeGen/trans_train.py
code served (permissive licence) · get_code("994c6a4e3b0b1a8d")
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 numpy as np

# Global rule list that the function uses
rule = [0, 1, 2, 3, 4, 5]

# Construct test arguments
words = ["apple", "^banana", "cherry", "^date", "elderberry", "fig"]
i = 4

ARGS = {
    "words": words,
    "i": i
}
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 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
find_father_index Ran
ran on a synthesized input
zysszy/TreeGen/trans_train.py
code served (permissive licence) · get_code("f328c6856bdf75bd")
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

torch.manual_seed(0)

# Construct arguments for find_father_index(words, i)
# The function takes a list of words and an index i
# It searches backwards through the list looking for a "father" word
# Words with '^' decrement a counter, others increment it
# Returns the index when counter reaches 1

words = [
    "apple",      # index 0
    "^banana",    # index 1 - has '^'
    "cherry",     # index 2
    "^date",      # index 3 - has '^'
    "elderberry", # index 4
    "^fig",       # index 5 - has '^'
    "grape"       # index 6
]

i = 4  # Start searching from index 4 ("elderberry")

ARGS = {
    "words": words,
    "i": i
}
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
findtheson Ran
ran on a synthesized input
zysszy/TreeGen/HS-B/eval/ast2code.py
code served (permissive licence) · get_code("6e8a00ebcf2fefb6")
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 int; 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

torch.manual_seed(0)

def findtheson(node):
    output = []
    count = 0
    site = 0
    for i in node[1:]:
        site += 1
        if i == '^':
            count -= 1
        else:
            count += 1
        if count < 0:
            break
        elif count == 1 and i != '^':
            output.append(site)
    return output

node = ['^', 'a', '^', 'b', 'c', '^', 'd']

ARGS = {
    'node': node
}
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 int — 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
get_deepth Ran
ran on a synthesized input
zysszy/TreeGen/trans_train.py
code served (permissive licence) · get_code("300dc56067ea2eaa")
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 numpy as np

# Global rule list (as referenced in the function)
rule = []

# Construct arguments for get_deepth function
# Parameters: line, rule, index_rule

# line: a string with words separated by spaces, some containing '^'
# The function parses this line and looks for 'node_gen' keyword
line = "parent_node child_node^ another^ node_gen grandchild^"

# rule: a list that gets modified during execution
# We'll pass the global rule (which starts empty)
rule_arg = []

# index_rule: an integer value used to update rule at a specific index
index_rule = 42

ARGS = {
    "line": line,
    "rule": rule_arg,
    "index_rule": index_rule,
}
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 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
line2vec Ran
ran on a synthesized input
zysszy/TreeGen/resolve_data.py
code served (permissive licence) · get_code("cb13989862d23ffd")
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 ndarray, ndim 1, dtype family float, shape (200,) 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 numpy as np

# Set up global vocabularies used by word2vec
vocabulary = {
    'hello': 10,
    'world': 20,
    'test': 30,
    'code': 40,
    'python': 50
}

tree_vocabulary = {
    'node': 100,
    'leaf': 110,
    'branch': 120,
    'root': 130,
    'child': 140
}

# Test data
line_input = "hello world test code python"
tp_input = "nl"
length_input = 200

ARGS = {
    'line': line_input,
    'tp': tp_input,
    'length': length_input
}
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 ndarray, ndim 1, dtype family float, shape (200,) — shape (200,) 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
tokenize_for_bleu_eval Ran
ran; our contract draft was wrong, not the code
zysszy/TreeGen/HS-B/eval/ast2code.py
pointer only (licence: MIT); the stored copy get_code serves, from yuxiang2/CoNaLa, is pointer only (licence: NONE) · get_code("59002967d7c8c2a8")
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 yuxiang2/CoNaLa, 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.

tqdm Ran
ran on a synthesized input · constructed an object; no result was computed
zysszy/TreeGen/resolve_data.py
code served (permissive licence) · get_code("a76095adda7142a8")
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, tqdm: 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

torch.manual_seed(0)

# The function tqdm(a) takes an iterable and wraps it with tqdm progress bar
# If tqdm is not available, it just returns the iterable as-is
# So we need to provide a valid iterable

# Create a simple iterable - a list works fine
iterable = [1, 2, 3, 4, 5]

ARGS = {
    "a": iterable
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tqdm — 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 tqdm 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
word2vec Ran
ran on a synthesized input
zysszy/TreeGen/resolve_data.py
code served (permissive licence) · get_code("6c2f0ab41920ac28")
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 numpy as np

torch.manual_seed(0)

# Set up the global vocabularies that the function uses
vocabulary = {
    'hello': 2,
    'world': 3,
    'python': 4,
    'code': 5
}

tree_vocabulary = {
    'node1': 10,
    'node2': 11,
    'subtree': 12,
    'leaf': 13
}

# Define ARGS mapping every parameter of word2vec to a value
ARGS = {
    'word': 'hello',
    'tp': 'nl'
}
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 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
adafactor_decay_rate_adam Not yet run
unverified (harvested, no recorded run)
zysszy/TreeGen/code_generate_model.py
code served (permissive licence) · get_code("7e84c28ab94718a8")
How it ran

no invocation recorded.

adafactor_decay_rate_pow Not yet run
unverified (harvested, no recorded run)
zysszy/TreeGen/code_generate_model.py
code served (permissive licence) · get_code("7a6613b9d1c8e6c8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 0 later) and 10 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
J_findtheson_name 1 other paper
1 dated earlier
1911.09983 (2019-11)
TreeGen: A Tree-Based Transformer Architecture for Code Generation
J_isend 1 other paper
1 dated earlier
1911.09983 (2019-11)
TreeGen: A Tree-Based Transformer Architecture for Code Generation
J_readlist 1 other paper
1 dated earlier
1911.09983 (2019-11)
TreeGen: A Tree-Based Transformer Architecture for Code Generation
tokenize_for_bleu_eval 1 other paper
1 dated earlier
1810.02720 (2018-10)
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation

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

Automated Program Repair (APR) helps improve the efficiency of software development and maintenance. Recent APR techniques use deep learning, particularly the encoder-decoder architecture, to generate patches. Though existing DL-based APR approaches have proposed different encoder architectures, the decoder remains to be the standard one, which generates a sequence of tokens one by one to replace the faulty statement. This decoder has multiple limitations: 1) allowing to generate syntactically incorrect programs, 2) inefficiently representing small edits, and 3) not being able to generate project-specific identifiers. In this paper, we propose Recoder, a syntax-guided edit decoder with placeholder generation. Recoder is novel in multiple aspects: 1) Recoder generates edits rather than modified code, allowing efficient representation of small edits; 2) Recoder is syntax-guided, with the novel provider/decider architecture to ensure the syntactic correctness of the patched program and accurate generation; 3) Recoder generates placeholders that could be instantiated as project-specific identifiers later. We conduct experiments to evaluate Recoder on 395 bugs from Defects4J v1.2 and 420 additional bugs from Defects4J v2.0. Our results show that Recoder repairs 53 bugs on Defects4J v1.2, which achieves 21.4% improvement over the previous state-of-the-art approach for single-hunk bugs (TBar). Importantly, to our knowledge, Recoder is the first DL-based APR approach that has outperformed the traditional APR approaches on this dataset. Furthermore, Recoder also repairs 19 bugs on the additional bugs from Defects4J v2.0, which is 137.5% more than TBar (8 bugs) and 850% more than SimFix (2 bugs). This result suggests that Recoder has better generalizability than existing APR approaches.

For agents

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

get_harvested_code_for_paper("2106.08253")
get_code_for_paper("2106.08253")
have("2106.08253")

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

curl https://syntology.ai/api/ran/2106.08253.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.08253.svg)](https://syntology.ai/paper/2106.08253)

Connect an agent — have() is free.