SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2601.09028 · 2026

OpenDecoder: Open Large Language Model Decoding to Incorporate Document Quality in RAG

Chao Zhang, Jinghan Zhang, Jian-Yun Nie, Fengran Mo, Zheyuan Liu, Tetsuya Sakai, Zhan Su, Jia Sun, Yuchen Hui

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 (0 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 10 failures of our instrument; 2 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/2601.09028.json · badge: /api/ran/2601.09028.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
fengranMark/OpenDecoder canonical (classifier label)
classifier basis: census
12 ran · 2 no recorded run
FunctionStatusWhere it lives
acc_score Ran
ran, honoured the contract we drafted
fengranMark/OpenDecoder/src/evaluation.py
code served (permissive licence) · get_code("baa96c8fcc9b7403")
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 json
import string
import re

def normalize_answer(s):
    """Lower text and remove punctuation, articles and extra whitespace."""

    def remove_articles(text):
        return re.sub('\\b(a|an|the)\\b', ' ', text)

    def white_space_fix(text):
        return ' '.join(text.split())

    def remove_punc(text):
        exclude = set(string.punctuation)
        return ''.join((ch for ch in text if ch not in exclude))

    def lower(text):
        return text.lower()
    return white_space_fix(remove_articles(remove_punc(lower(s))))

def acc_score(prediction, ground_truth):
    prediction = normalize_answer(prediction)
    ground_truth = normalize_answer(ground_truth)
    if ground_truth in prediction:
        return 1
    return 0

ARGS = {
    "prediction": "The quick brown fox jumps over the lazy dog.",
    "ground_truth": "quick brown fox"
}
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 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
extract_first_digit Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/retrieval/generate_QPP_scores.py
code served (permissive licence) · get_code("a516d4163636a590")
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

torch.manual_seed(0)

def extract_first_digit(text):
    for char in text:
        if char.isdigit():
            return char
    return '0'

ARGS = {
    'text': 'hello 123 world'
}
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
f1_score Ran
ran, honoured the contract we drafted
fengranMark/OpenDecoder/src/evaluation.py
code served (permissive licence) · get_code("e08cad149be1d1ab")
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 float, 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.

import string
import re
from collections import Counter

# The function f1_score takes two string parameters: prediction and ground_truth
# It normalizes both strings and computes F1 score based on token overlap

ARGS = {
    "prediction": "The quick brown fox jumps over the lazy dog.",
    "ground_truth": "A quick brown fox jumps over a lazy dog"
}
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 float, 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
load_imodel_and_iconfig_package Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/inference.py
code served (permissive licence) · get_code("ae944771f0673fa8")
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 os
import sys
import tempfile
import importlib.util

# Create a temporary directory structure to simulate a valid model package
temp_dir = tempfile.mkdtemp()
src_path = temp_dir

# Create the model directory
model_dir = os.path.join(src_path, 'model')
os.makedirs(model_dir, exist_ok=True)

# Create a test model package directory
model_pattern = 'test_model'
package_dir = os.path.join(model_dir, model_pattern)
os.makedirs(package_dir, exist_ok=True)

# Create __init__.py for the package
with open(os.path.join(package_dir, '__init__.py'), 'w') as f:
    f.write('')

# Create configuration.py with IConfig class
with open(os.path.join(package_dir, 'configuration.py'), 'w') as f:
    f.write('''
class IConfig:
    def __init__(self):
        self.hidden_size = 768
        self.num_layers = 12
''')

# Create modeling.py with IModelForCausalLM class
with open(os.path.join(package_dir, 'modeling.py'), 'w') as f:
    f.write('''
class IModelForCausalLM:
    def __init__(self, config):
        self.config = config
        self.hidden_size = config.hidden_size
''')

# Define ARGS with valid parameters
ARGS = {
    'model_pattern': 'test_model',
    'src_path': src_path
}
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
load_imodel_and_iconfig_package Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/train_parallel.py
code served (permissive licence) · get_code("2a83dab612562e4d")
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 os
import sys
import tempfile
import importlib.util

# Create a temporary directory structure for the model package
temp_dir = tempfile.mkdtemp()
src_path = temp_dir

# Create the model directory
model_dir = os.path.join(src_path, 'model')
os.makedirs(model_dir, exist_ok=True)

# Create a test model pattern name
model_pattern = 'test_model'
model_package_dir = os.path.join(model_dir, model_pattern)
os.makedirs(model_package_dir, exist_ok=True)

# Create __init__.py files
with open(os.path.join(model_package_dir, '__init__.py'), 'w') as f:
    f.write('')

# Create modeling.py with IModelForCausalLM class
modeling_code = '''
class IModelForCausalLM:
    def __init__(self):
        self.name = "test_model"
'''

with open(os.path.join(model_package_dir, 'modeling.py'), 'w') as f:
    f.write(modeling_code)

# Create configuration.py with IConfig class
configuration_code = '''
class IConfig:
    def __init__(self):
        self.hidden_size = 768
'''

with open(os.path.join(model_package_dir, 'configuration.py'), 'w') as f:
    f.write(configuration_code)

# Define the arguments for the function
ARGS = {
    'model_pattern': 'test_model',
    'src_path': src_path
}
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
load_imodel_and_iconfig_package Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/utils/iniModel.py
code served (permissive licence) · get_code("8033a98192281151")
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 os
import sys
import tempfile
import shutil

# Create a temporary directory structure to simulate a valid model package
temp_dir = tempfile.mkdtemp()
src_path = temp_dir

# Create the model directory
model_path = os.path.join(src_path, 'model')
os.makedirs(model_path, exist_ok=True)

# Create a sample model pattern directory
model_pattern = 'test_model'
model_pattern_path = os.path.join(model_path, model_pattern)
os.makedirs(model_pattern_path, exist_ok=True)

# Create __init__.py files
with open(os.path.join(model_pattern_path, '__init__.py'), 'w') as f:
    f.write('')

# Create modeling.py with IModelForCausalLM class
modeling_code = '''
class IModelForCausalLM:
    def __init__(self):
        pass
'''
with open(os.path.join(model_pattern_path, 'modeling.py'), 'w') as f:
    f.write(modeling_code)

# Create configuration.py with IConfig class
configuration_code = '''
class IConfig:
    def __init__(self):
        pass
'''
with open(os.path.join(model_pattern_path, 'configuration.py'), 'w') as f:
    f.write(configuration_code)

ARGS = {
    'model_pattern': 'test_model',
    'src_path': src_path
}
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
normalize_answer Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/evaluation.py
code served (permissive licence) · get_code("24a31436a8dbd623")
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 json
import string
import re

def normalize_answer(s):
    """Lower text and remove punctuation, articles and extra whitespace."""

    def remove_articles(text):
        return re.sub('\\b(a|an|the)\\b', ' ', text)

    def white_space_fix(text):
        return ' '.join(text.split())

    def remove_punc(text):
        exclude = set(string.punctuation)
        return ''.join((ch for ch in text if ch not in exclude))

    def lower(text):
        return text.lower()
    return white_space_fix(remove_articles(remove_punc(lower(s))))


# The function takes a single parameter 's' which should be a string
ARGS = {
    "s": "The Quick Brown Fox! Is jumping over a lazy dog."
}
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
padding_seq_to_same_length Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
fengranMark/OpenDecoder/src/retrieval/retrieval_dataset.py
pointer only (licence: MIT); the stored copy get_code serves, from fengranMark/QRACDR, is pointer only (licence: NONE) · get_code("58745a56a67a45ad")
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 fengranMark/QRACDR, 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.

parser_binary Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/retrieval/generate_QPP_scores.py
code served (permissive licence) · get_code("7f3a7ddcef1209a7")
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

torch.manual_seed(0)

def parser_binary(text):
    if text in ['Relevant', 'Irrelevant']:
        return '1' if text == 'Relevant' else '0'
    print(f'Parsing:***********\nOriginal text:\n{text}\n')
    if 'Relevant' in text:
        text_ = '1'
    elif 'Irrelevant' in text or 'Ir' in text:
        text_ = '0'
    else:
        text_ = '0'
    print(f'Parsed text:\n{text_}\n***********\n')
    return text_

ARGS = {
    'text': 'Relevant'
}
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
parser_digit Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/retrieval/generate_QPP_scores.py
code served (permissive licence) · get_code("aa8d783be138ff41")
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

torch.manual_seed(0)

def extract_first_digit(text):
    for char in text:
        if char.isdigit():
            return char
    return '0'

def parser_digit(text):
    if text in ['0', '1', '2', '3', '4']:
        return text
    else:
        print(f'Parsing:***********\nOriginal text:\n{text}\n')
        text_ = extract_first_digit(text)
        print(f'Parsed text:\n{text_}\n***********\n')
    return text_

ARGS = {
    'text': 'abc123def'
}
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
pload Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/RAG_dataset.py
code served (permissive licence) · get_code("aa4b9685f084f5f8")
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 pickle
import tempfile
import os

# Create a temporary pickle file with sample data
temp_dir = tempfile.mkdtemp()
temp_path = os.path.join(temp_dir, "test_data.pkl")

sample_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "dict"}
}

with open(temp_path, 'wb') as f:
    pickle.dump(sample_data, f)

ARGS = {
    "path": temp_path
}
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
pload Ran
ran; our contract draft was wrong, not the code
fengranMark/OpenDecoder/src/retrieval/construct_indicators.py
code served (permissive licence) · get_code("155939bd2bae6551")
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 pickle
import tempfile
import os

# Create a temporary pickle file for testing
temp_dir = tempfile.mkdtemp()
test_file = os.path.join(temp_dir, "test_object.pkl")

# Create a simple test object and pickle it
test_object = {
    "data": [1, 2, 3, 4, 5],
    "name": "test",
    "nested": {"key": "value"}
}

with open(test_file, 'wb') as f:
    pickle.dump(test_object, f)

ARGS = {
    "path": test_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 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
calculate_similarity Not yet run
unverified (harvested, no recorded run)
fengranMark/OpenDecoder/src/retrieval/construct_indicators.py
code served (permissive licence) · get_code("9ba9e62b273b7c2b")
How it ran

no invocation recorded.

get_embedding Not yet run
unverified (harvested, no recorded run)
fengranMark/OpenDecoder/src/retrieval/construct_indicators.py
code served (permissive licence) · get_code("0bbe88e1e97834eb")
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, 7 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 1 the same month, 3 later) and 7 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
normalize_answer 133 other papers
116 dated earlier, 3 the same month, 13 later, 1 undated · 1 only as an assembly brick
1508.06615 (2015-08)
Character-Aware Neural Language Models
1606.05250 (2016-06)
SQuAD: 100,000+ Questions for Machine Comprehension of Text
1608.07905 (2016-08)
Machine Comprehension Using Match-LSTM and Answer Pointer
and 130 more not named here
f1_score 41 other papers
41 dated earlier
1508.06615 (2015-08)
Character-Aware Neural Language Models
1606.05250 (2016-06)
SQuAD: 100,000+ Questions for Machine Comprehension of Text
1608.07905 (2016-08)
Machine Comprehension Using Match-LSTM and Answer Pointer
and 38 more not named here
extract_first_digit 1 other paper
1 dated earlier
2404.01012 (2024-04)
Query Performance Prediction using Relevance Judgments Generated by Large Language Models
padding_seq_to_same_length 1 other paper
1 later
2607.08540 (2026-07)
Improving Ad-hoc Search Effectiveness for Conversational Information Retrieval via Model Merging
parser_binary 1 other paper
1 dated earlier
2404.01012 (2024-04)
Query Performance Prediction using Relevance Judgments Generated by Large Language Models
parser_digit 1 other paper
1 dated earlier
2404.01012 (2024-04)
Query Performance Prediction using Relevance Judgments Generated by Large Language Models
pload 1 other paper
1 later
2607.08540 (2026-07)
Improving Ad-hoc Search Effectiveness for Conversational Information Retrieval via Model Merging

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

The development of large language models (LLMs) has achieved superior performance in a range of downstream tasks, including LLM-based retrieval-augmented generation (RAG). The quality of generated content heavily relies on the usefulness of the retrieved information and the capacity of LLMs' internal information processing mechanism to incorporate it in answer generation. It is generally assumed that the retrieved information is relevant to the question. However, the retrieved information may have a variable degree of relevance and usefulness, depending on the question and the document collection. It is important to take into account the relevance of the retrieved information in answer generation. In this paper, we propose OpenDecoder, a new approach that leverages explicit evaluation of the retrieved information as quality indicator features for generation. We aim to build a RAG model that is more robust to varying levels of noisy context. Three types of explicit evaluation information are considered: relevance score, ranking score, and QPP (query performance prediction) score. The experimental results on five benchmark datasets demonstrate the effectiveness and better robustness of OpenDecoder by outperforming various baseline methods. Importantly, this paradigm is flexible to be integrated with the post-training of LLMs for any purposes and incorporated with any type of external indicators.

For agents

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

get_harvested_code_for_paper("2601.09028")
get_code_for_paper("2601.09028")
have("2601.09028")

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

curl https://syntology.ai/api/ran/2601.09028.json

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

[![Syntology run record](https://syntology.ai/api/ran/2601.09028.svg)](https://syntology.ai/paper/2601.09028)

Connect an agent — have() is free.