SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2510.00307 · ICLR · 2025

BiasBusters: Uncovering and Mitigating Tool Selection Bias in Large Language Models

Philip Torr, Yarin Gal, Adel Bibi, Jialin Yu, Zixuan Li, Alasdair Paren, Thierry Blankenstein, Vassilis Plachouras, Sunando Sengupta

arXiv · PDF · Open in the Atlas

Code that ran

10 of 12 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; 4 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 6 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/2510.00307.json · badge: /api/ran/2510.00307.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
thierry123454/tool-selection-bias canonical (classifier label)
classifier basis: census
10 ran · 2 no recorded run
FunctionStatusWhere it lives
approx_tokens Ran
ran, honoured the contract we drafted
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/generate_data_gemini.py
pointer only (licence: NONE) · get_code("9100f6fca4585d96")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

compute_stats Ran
ran, honoured the contract we drafted
thierry123454/tool-selection-bias/5_bias_investigation/analyze_features/calc_stats_feats.py
pointer only (licence: NONE) · get_code("93a14e5a9ccc1a74")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

cosine_similarity Ran
ran; our contract draft was wrong, not the code
thierry123454/tool-selection-bias/5_bias_investigation/analyze_features/sense_of_scale.py
pointer only (licence: NONE) · get_code("2bad52a890367fbe")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

count_existing_lines Ran
ran, honoured the contract we drafted
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/generate_data_gemini.py
pointer only (licence: NONE) · get_code("ff4ecc06bb7d6319")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

count_words Ran
ran, honoured the contract we drafted
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/generate_data_gemini.py
pointer only (licence: NONE) · get_code("2498e76b5f8a7bd9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

escape_tex Ran
ran; our contract draft was wrong, not the code
thierry123454/tool-selection-bias/5_bias_investigation/analyze_features/lr.py
pointer only (licence: NONE) · get_code("7d76e26c69b1b999")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

example_call Ran
ran; our contract draft was wrong, not the code
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/generate_data.py
pointer only (licence: NONE) · get_code("58cafcae23a256be")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

load_json Ran
ran; our contract draft was wrong, not the code
thierry123454/tool-selection-bias/5_bias_investigation/extract_features/add_selection_rate.py
pointer only (licence: NONE) · get_code("a0c7393409fd0be9")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

render_param Ran
ran; our contract draft was wrong, not the code
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/generate_data.py
pointer only (licence: NONE) · get_code("64903b61e4b03492")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

standardize Ran
ran; our contract draft was wrong, not the code
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/plot_compare_base_vs_adapted.py
pointer only (licence: NONE) · get_code("853c161b9ef4dc1e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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.

get_conv_template Not yet run
unverified (harvested, no recorded run)
thierry123454/tool-selection-bias/toolbench/tool_conversation.py
pointer only (licence: NONE) · get_code("eab40e1435b85709")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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. No invocation recorded.

parse_action Not yet run
unverified (harvested, no recorded run)
thierry123454/tool-selection-bias/5_bias_investigation/continued_pretraining/plot_compare_base_vs_adapted.py
pointer only (licence: NONE) · get_code("98bbe08b08d7a02b")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, thierry123454/tool-selection-bias 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. No invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 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, 2 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
get_conv_template 69 other papers
57 dated earlier, 8 later, 3 same year, order not determinable, 1 undated
1911.03705 (2019-11)
CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning
2112.05682 (2021-12)
Self-attention Does Not Need $O(n^2)$ Memory
2204.07705 (2022-04)
SUPER-NATURALINSTRUCTIONS: Generalization via Declarative Instructions on 1600+ NLP Tasks
and 66 more not named here
load_json 7 other papers
5 dated earlier, 2 later
2411.10954 (2024-11)
Dialectal Toxicity Detection: Evaluating LLM-as-a-Judge Consistency Across Language Varieties
2504.04907 (2025-04)
Video-Bench: Human-Aligned Video Generation Benchmark
2506.19054 (2025-06)
POLY-GUARD: Multi-Domain, Policy-Grounded, AI Security Guardrail Benchmark
and 4 more not named here

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

Agents backed by large language models (LLMs) increasingly rely on external tools drawn from marketplaces where multiple providers offer functionally equivalent options. This raises a critical fairness concern: systematic bias in tool selection can degrade user experience and distort competition by privileging certain providers over others. We introduce a benchmark of diverse tool categories, each containing multiple functionally equivalent tools, to systematically evaluate toolselection bias. Using this benchmark, we evaluate seven LLMs and show that substantial bias persists, with models either fixating on a single provider or disproportionately favoring tools that appear earlier in the context. To uncover the sources of this behavior, we conduct controlled experiments that isolate the effects of tool features, exposed metadata (name, description, and parameters), and pre-training exposure. We find that (1) semantic alignment between user queries and tool metadata is the strongest driver of selection; (2) small perturbations to tool descriptions can significantly shift choices; and (3) repeated pre-training exposure to a single endpoint amplifies provider-level bias. Finally, we propose a lightweight mitigation strategy that first filters tools to a relevant subset and then samples uniformly, substantially reducing selection bias while maintaining strong task coverage. Our results highlight tool-selection bias as a key obstacle to the fair deployment of tool-augmented LLM agents. Our code and benchmark are publicly available at https://github.com/thierry123454/tool-selection-bias.

For agents

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

get_harvested_code_for_paper("2510.00307")
get_code_for_paper("2510.00307")
have("2510.00307")

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

curl https://syntology.ai/api/ran/2510.00307.json

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

[![Syntology run record](https://syntology.ai/api/ran/2510.00307.svg)](https://syntology.ai/paper/2510.00307)

Connect an agent — have() is free.