SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.08845 · 2023

Large Language Models are Zero-Shot Rankers for Recommender Systems

arXiv · PDF · Open in the Atlas

Code that ran

4 of 4 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; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 0 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/2305.08845.json · badge: /api/ran/2305.08845.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
rucaibox/llmrank canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
FunctionStatusWhere it lives
get_user2count Ran
ran on a synthesized input
rucaibox/llmrank/llmrank/data_process_amazon.py
code served (permissive licence) · get_code("dd88baa2e39b7ec5")
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 defaultdict. 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

# Analyze the function: it takes `inters` which is iterable of units,
# where each unit is indexable (unit[0] accesses first element)
# The function counts occurrences of unit[0] values

# Create a list of tuples/lists where first element is a user ID
inters = [
    (1, 'action_a'),
    (2, 'action_b'),
    (1, 'action_c'),
    (3, 'action_d'),
    (2, 'action_e'),
]

ARGS = {
    'inters': inters,
}
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 defaultdict — a defaultdict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a defaultdict 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
load_meta_items Ran
ran on a synthesized input
rucaibox/llmrank/llmrank/data_process_amazon.py
code served (permissive licence) · get_code("423c5d0056286c09")
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 gzip
import json
import tempfile
import os

# Create a temporary gzip file with valid metadata
temp_file = tempfile.NamedTemporaryFile(mode='wb', suffix='.gz', delete=False)
temp_path = temp_file.name

# Write sample metadata in the expected format
sample_data = [
    {
        'asin': 'B001',
        'title': 'Product One',
        'brand': 'Brand A',
        'category': ['Electronics', 'Gadgets']
    },
    {
        'asin': 'B002',
        'title': 'Product\tTwo',  # Contains tab to test replacement
        'brand': 'Brand B',
        'category': ['Home']
    },
    {
        'asin': 'B003',
        'title': '',  # Empty title - should be skipped
        'brand': 'Brand C',
        'category': ['Sports']
    },
    {
        'asin': 'B004',
        'title': 'Product Four',
        'brand': 'Brand D',
        'category': ['Books', 'Fiction']
    }
]

with gzip.open(temp_path, 'wb') as f:
    for item in sample_data:
        f.write((json.dumps(item) + '\n').encode('utf-8'))

temp_file.close()

ARGS = {
    'file': temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tqdm — 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
load_ratings Ran
ran on a synthesized input
rucaibox/llmrank/llmrank/data_process_amazon.py
code served (permissive licence) · get_code("ffcd7499f55eb794")
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 tempfile
import os

# Create a temporary file with sample rating data
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.csv')
temp_file.write("item1,user1,5,1000\n")
temp_file.write("item2,user2,4,1001\n")
temp_file.write("item3,user1,3,1002\n")
temp_file.write("item1,user3,5,1003\n")
temp_file.close()

ARGS = {
    'file': temp_file.name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tqdm — 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_single_source Ran
ran on a synthesized input
rucaibox/llmrank/llmrank/mix_sources.py
code served (permissive licence) · get_code("38d0b46943ff2ffc")
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 os
import tempfile

# Create a temporary directory and test data file
temp_dir = tempfile.mkdtemp()

# Create a sample data file with the expected format
dataset_name = "test_dataset"
source_name = "source1"
filepath = os.path.join(temp_dir, f"{dataset_name}.{source_name}")

with open(filepath, 'w', encoding='utf-8') as f:
    f.write("user1\titem1 item2 item3\n")
    f.write("user2\titem4 item5\n")
    f.write("user3\titem6 item7 item8 item9\n")

config = {
    'data_path': temp_dir,
    'dataset': dataset_name
}

ARGS = {
    'config': config,
    'source_name': source_name
}
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 — 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

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 1 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
get_user2count 2 other papers
1 dated earlier, 1 later
2206.05941 (2022-06)
Towards Universal Sequence Representation Learning for Recommender Systems
2025.findings-acl.307 (2025 (year only))
Expectation Confirmation Preference Optimization for Multi-Turn Conversational Recommendation Agent
load_ratings 1 other paper
1 dated earlier
2206.05941 (2022-06)
Towards Universal Sequence Representation Learning for Recommender Systems

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

Recently, large language models (LLMs) (e.g., GPT-4) have demonstrated impressive general-purpose task-solving abilities, including the potential to approach recommendation tasks. Along this line of research, this work aims to investigate the capacity of LLMs that act as the ranking model for recommender systems. We first formalize the recommendation problem as a conditional ranking task, considering sequential interaction histories as conditions and the items retrieved by other candidate generation models as candidates. To solve the ranking task by LLMs, we carefully design the prompting template and conduct extensive experiments on two widely-used datasets. We show that LLMs have promising zero-shot ranking abilities but (1) struggle to perceive the order of historical interactions, and (2) can be biased by popularity or item positions in the prompts. We demonstrate that these issues can be alleviated using specially designed prompting and bootstrapping strategies. Equipped with these insights, zero-shot LLMs can even challenge conventional recommendation models when ranking candidates are retrieved by multiple candidate generators. The code and processed datasets are available at https://github.com/RUCAIBox/LLMRank.

For agents

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

get_harvested_code_for_paper("2305.08845")
get_code_for_paper("2305.08845")
have("2305.08845")

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

curl https://syntology.ai/api/ran/2305.08845.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.08845.svg)](https://syntology.ai/paper/2305.08845)

Connect an agent — have() is free.