SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2605.09236 · 2026

Matching Meaning at Scale: Evaluating Semantic Search for 18th-Century Intellectual History through the Case of Locke

Yu Wu, Mikko Tolonen, Filip Ginter, Michael Mathioudakis, Ananth Mahadevan

arXiv · PDF

Code that ran

13 of 14 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; 3 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 9 failures of our instrument; 1 no recorded run), measured 2026-11-22; 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/2605.09236.json · badge: /api/ran/2605.09236.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
TurkuNLP/semantic-sim canonical (classifier label)
classifier basis: census
13 ran · 1 no recorded run
FunctionStatusWhere it lives
collate Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/embed_data.py
code served (permissive licence) · get_code("6ed105130bd872bf")
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

torch.manual_seed(0)

# Create sample dataset items that match what collate expects
itemlist = [
    {
        'enc': [1, 2, 3],
        'attention_mask': [1, 1, 1],
        'spec_token_mask': [0, 0, 1],
        'token_type_id': [0, 0, 0],
        'line_idx': 0,
        'txt': 'sample text 1'
    },
    {
        'enc': [4, 5],
        'attention_mask': [1, 1],
        'spec_token_mask': [1, 0],
        'token_type_id': [0, 1],
        'line_idx': 1,
        'txt': 'sample text 2'
    }
]

ARGS = {
    'itemlist': itemlist
}
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 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
embed_batch Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
TurkuNLP/semantic-sim/embed.py
code served (permissive licence) · get_code("c97754bcbf6fdea1")
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. 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
import torch.nn as nn

torch.manual_seed(0)

# Create a minimal BERT-like model that returns last_hidden_state
class MinimalBertModel(nn.Module):
    def __init__(self, hidden_size=4):
        super().__init__()
        self.hidden_size = hidden_size
        self.linear = nn.Linear(hidden_size, hidden_size)
    
    def forward(self, input_ids, attention_mask, token_type_ids):
        # Return a minimal output with last_hidden_state attribute
        batch_size = input_ids.shape[0]
        seq_length = input_ids.shape[1]
        last_hidden_state = torch.randn(batch_size, seq_length, self.hidden_size)
        
        # Create a simple namespace-like object to hold last_hidden_state
        class Output:
            pass
        
        output = Output()
        output.last_hidden_state = last_hidden_state
        return output

# Create a small batch with required keys
batch_size = 2
seq_length = 5
hidden_dim = 4

batch = {
    'enc': torch.randint(0, 1000, (batch_size, seq_length)),
    'attention_mask': torch.ones(batch_size, seq_length, dtype=torch.long),
    'token_type_id': torch.zeros(batch_size, seq_length, dtype=torch.long),
    'spec_token_mask': torch.zeros(batch_size, seq_length, dtype=torch.long),
}

# Create model
bert_model = MinimalBertModel(hidden_size=hidden_dim)

ARGS = {
    'batch': batch,
    'bert_model': bert_model,
}
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
embed_batch Ran
ran, honoured the contract we drafted
TurkuNLP/semantic-sim/embed_sbert.py
code served (permissive licence) · get_code("ffa9d49520a1a132")
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 numpy as np

torch.manual_seed(0)

class MockSBERTModel:
    def encode(self, texts):
        # Return numpy array of embeddings
        # texts can be a string or list of strings
        if isinstance(texts, str):
            texts = [texts]
        # Return embeddings of shape (batch_size, embedding_dim)
        return np.random.randn(len(texts), 4).astype(np.float32)

batch = {
    'txt': ['hello world', 'test sentence']
}

sbert_model = MockSBERTModel()

ARGS = {
    'batch': batch,
    'sbert_model': sbert_model
}
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 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
grouper Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/interactive_index.py
code served (permissive licence) · get_code("2d0224d483d91857")
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 zip_longest. 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 itertools

def grouper(iterable, n, fillvalue=None):
    """Collect data into non-overlapping fixed-length chunks or blocks"""
    args = [iter(iterable)] * n
    return itertools.zip_longest(*args, fillvalue=fillvalue)

# Setup code
ARGS = {
    "iterable": [1, 2, 3, 4, 5, 6, 7, 8],
    "n": 3,
    "fillvalue": None
}
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 zip_longest — a zip_longest output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a zip_longest 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
line_about_ok Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/clean_data.py
code served (permissive licence) · get_code("c5b52464dd6c62a5")
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 line_about_ok(s):
    n_rough_words = len(s.split())
    if n_rough_words < 3 or n_rough_words > 50:
        return (False, 'too few words')
    n_punct_and_junk = len(re.sub('[a-zA-ZäöåÅÄÖ0-9:"`\'(""\',.!?() -]+', '', s))
    if n_punct_and_junk / len(s) > 0.1:
        return (False, 'too many punct')
    n_num = len(re.sub('[^0-9]+', ' ', s).split())
    if n_num > 5 and n_num / len(s) > 0.2:
        return (False, 'too many nums')
    return (True, None)

# Construct test arguments
ARGS = {
    's': 'The quick brown fox jumps over the lazy dog today'
}
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
line_clean Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/clean_data.py
code served (permissive licence) · get_code("24df7eb1a19e250b")
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 remove_initial_dash(s):
    return re.sub('^\\s*-+\\s*', '', s)

def line_about_ok(s):
    n_rough_words = len(s.split())
    if n_rough_words < 3 or n_rough_words > 50:
        return (False, 'too few words')
    n_punct_and_junk = len(re.sub('[a-zA-ZäöåÅÄÖ0-9:"`\'(""\',.!?() -]+', '', s))
    if n_punct_and_junk / len(s) > 0.1:
        return (False, 'too many punct')
    n_num = len(re.sub('[^0-9]+', ' ', s).split())
    if n_num > 5 and n_num / len(s) > 0.2:
        return (False, 'too many nums')
    return (True, None)

def line_clean(line):
    line = line.strip()
    if not line:
        return None
    line = remove_initial_dash(line)
    ok, reason = line_about_ok(line)
    if ok:
        return (True, 'ok', line.strip())
    else:
        return (False, reason, line.strip())

ARGS = {
    'line': 'This is a valid test line with enough words for processing'
}
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
main_cls Ran
ran on a synthesized input
TurkuNLP/semantic-sim/plot.py
code served (permissive licence) · get_code("7979f5424cb0defe")
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
import numpy as np

torch.manual_seed(0)

ARGS = {
    'cls': '1<'
}
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 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
open_possibly_gz_file Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/id2txt.py
code served (permissive licence) · get_code("43c3943e2bfa8ea9")
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 TextIOWrapper. 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 gzip
import os
import tempfile

# Create a temporary directory and files for testing
temp_dir = tempfile.mkdtemp()

# Create a plain text file
plain_file = os.path.join(temp_dir, "test.txt")
with open(plain_file, 'w') as f:
    f.write("test content")

# Create a gzipped file
gz_file = os.path.join(temp_dir, "test.gz")
with gzip.open(gz_file, 'wt') as f:
    f.write("gzipped content")

# Create a file that exists as both plain and .gz version
plain_dual = os.path.join(temp_dir, "dual")
with open(plain_dual, 'w') as f:
    f.write("dual plain")

# Test cases: None, string path to .gz file, string path to plain file,
# string path where .gz exists, and a file-like object
import io
file_like = io.StringIO("file-like content")

ARGS = {
    "f_or_name": plain_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 TextIOWrapper — a TextIOWrapper 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
pad_with_zero Ran
ran, honoured the contract we drafted
TurkuNLP/semantic-sim/embed_data.py
code served (permissive licence) · get_code("b15b3e24b6e9ffbd")
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 int, shape (3, 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)

# pad_with_zero takes a single parameter 'vals' which is expected to be
# an iterable of sequences that can be converted to LongTensors
# The function converts each to LongTensor and pads them to the same length

# Create a list of lists with different lengths to demonstrate padding
vals = [
    [1, 2, 3],
    [4, 5],
    [6, 7, 8, 9]
]

ARGS = {
    "vals": vals
}
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 int, shape (3, 4) — shape (3, 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
random_sample_of_batches Ran
ran, honoured the contract we drafted
TurkuNLP/semantic-sim/create_faiss_index.py
code served (permissive licence) · get_code("61a1aa1e32f835b2")
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 (8, 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 tempfile
import pickle
import os

torch.manual_seed(0)

# Create temporary batch files with pickled data
temp_dir = tempfile.mkdtemp()
batch_files = []

# Create 3 small batch files
for i in range(3):
    batch_file = os.path.join(temp_dir, f"batch_{i}.pkl")
    batch_files.append(batch_file)
    
    with open(batch_file, 'wb') as f:
        # Write 2 batches per file
        for j in range(2):
            sent_idx = i * 2 + j
            embedding_batch = torch.randn(2, 4)  # batch_size=2, embedding_dim=4
            pickle.dump((sent_idx, embedding_batch), f)

ARGS = {
    'batch_files': batch_files,
    'proportion': 0.5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — 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 (8, 4) — shape (8, 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
read_results Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/plot.py
code served (permissive licence) · get_code("cc60092f64f71a73")
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 collections

def read_results(inp):
    counter = {}
    for line in inp:
        line = line.strip()
        cls, position, sent_L, sent_R = line.split('\t')
        if cls.startswith('4>') and '-rev' in cls:
            continue
        if cls.startswith('4<') and '-rev' not in cls:
            continue
        if position == 'None':
            position = None
        else:
            position = int(position)
        coarse_cls = main_cls(cls)
        if position is None:
            pos = 'NA'
        else:
            for cutoff in (1, 10, 100, 1000, 2048):
                if position - 1 < cutoff:
                    pos = str(cutoff)
                    break
            else:
                assert False
        counter.setdefault(coarse_cls, []).append(pos)
    for k, pos_list in counter.items():
        counter[k] = collections.Counter(pos_list)
    result = {}
    for k, pos_list in counter.items():
        tot = sum((v for k, v in pos_list.items()))
        res_list = []
        for r in ('1', '10', '100', '1000', '2048', 'NA'):
            res_list.append(pos_list[r] / tot * 100)
        result[k] = res_list
    return result

def main_cls(cls):
    coarse = cls[0]
    assert coarse.isnumeric(), cls
    if len(cls) > 1 and cls[1] in ('<', '>'):
        coarse += '<>'
    return coarse

inp_lines = [
    "1\t5\tleft_sent\tright_sent",
    "2>\t50\tleft_sent\tright_sent",
    "3<\t500\tleft_sent\tright_sent",
    "4>\t1500\tleft_sent\tright_sent",
    "4<-rev\t100\tleft_sent\tright_sent",
    "5\tNone\tleft_sent\tright_sent",
]

ARGS = {
    "inp": inp_lines
}
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 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
remove_initial_dash Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/clean_data.py
code served (permissive licence) · get_code("77382c479342ef23")
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 re

def remove_initial_dash(s):
    return re.sub('^\\s*-+\\s*', '', s)

ARGS = {
    's': '   --- hello world'
}
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 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
s2hash Ran
ran; our contract draft was wrong, not the code
TurkuNLP/semantic-sim/index_sentences.py
code served (permissive licence) · get_code("0a1348b9e265f50a")
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 hashlib

def s2hash(idx_sent):
    idx, sent = idx_sent
    return (idx, hashlib.sha1(sent.encode('utf-8')).hexdigest()[:15])

# The function expects a tuple (idx, sent) where:
# - idx: an index (integer or similar)
# - sent: a string that will be encoded to UTF-8 and hashed

ARGS = {
    'idx_sent': (0, "hello world")
}
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
open_possibly_gz_file Not yet run
unverified (harvested, no recorded run)
TurkuNLP/semantic-sim/interactive_index.py
code served (permissive licence) · get_code("033b3990e98095d9")
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 13 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 13 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

While digitized corpora have transformed the study of intellectual transmission, current methods rely heavily on lexical text reuse detection, capturing verbatim quotations but fundamentally missing paraphrases and complex implicit engagement. This paper evaluates semantic search in 18th-century intellectual history through the reception of John Locke's foundational work. Using expert annotation grounded in a semantic taxonomy, we examine whether an off-the-shelf semantic search pipeline can surface meaning-level correspondences overlooked by lexical methods. Our results demonstrate that semantic search retrieves substantially more implicit receptions than lexical baselines. However, linguistic diagnostics also reveal a "lexical gatekeeping" effect, where retrieval remains partially constrained by surface vocabulary overlap. These findings highlight both the potential and the limitations of semantic retrieval for analyzing the circulation of ideas in large historical corpora. The data is available at https: //github.com/COMHIS/locke-sim-data.

For agents

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

get_harvested_code_for_paper("2605.09236")
get_code_for_paper("2605.09236")
have("2605.09236")

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

curl https://syntology.ai/api/ran/2605.09236.json

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

[![Syntology run record](https://syntology.ai/api/ran/2605.09236.svg)](https://syntology.ai/paper/2605.09236)

Connect an agent — have() is free.