@inproceedings{acikgoz-etal-2026-tool,
title = "{T}ool-{R}0: Self-Evolving {LLM} Agents for Tool-Learning from Zero Data",
author = {Acikgoz, Emre Can and
Qian, Cheng and
H{\"u}botter, Jonas and
Ji, Heng and
Hakkani-T{\"u}r, Dilek and
Tur, Gokhan},
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1407/",
pages = "29989--30017",
ISBN = "979-8-89176-462-0",
abstract = "Large language models (LLMs) are becoming the foundation for autonomous agents that can use tools to solve complex tasks. Reinforcement learning (RL) has emerged as a common approach for injecting such agentic capabilities, but typically under tightly controlled training setups. It often depends on carefully curated task{--}solution pairs and substantial human supervision, which creates a fundamental obstacle to open-ended self-evolution. In this paper, we propose Tool-R0 for training general purpose tool-calling agents from scratch with self-play RL, under a zero-data assumption. Initialized from the same base LLM, Tool-R0 co-evolves a Generator and a Solver with complementary rewards: one proposes targeted challenging tasks at the other{'}s competence frontier and the other learns to solve them with real-world tool calls. This creates a self-evolving cycle that requires no pre-existing datasets. Evaluation on five tool-use benchmarks show that Tool-R0 yields 92.5{\%} relative improvement over the base model and surpasses fully supervised tool-calling baselines under the same setting. Our work further provides empirical insights into self-play LLM agents by analyzing co-evolution, curriculum dynamics, and scaling behavior."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="acikgoz-etal-2026-tool">
<titleInfo>
<title>Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero Data</title>
</titleInfo>
<name type="personal">
<namePart type="given">Emre</namePart>
<namePart type="given">Can</namePart>
<namePart type="family">Acikgoz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Cheng</namePart>
<namePart type="family">Qian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jonas</namePart>
<namePart type="family">Hübotter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Heng</namePart>
<namePart type="family">Ji</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dilek</namePart>
<namePart type="family">Hakkani-Tür</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gokhan</namePart>
<namePart type="family">Tur</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Large language models (LLMs) are becoming the foundation for autonomous agents that can use tools to solve complex tasks. Reinforcement learning (RL) has emerged as a common approach for injecting such agentic capabilities, but typically under tightly controlled training setups. It often depends on carefully curated task–solution pairs and substantial human supervision, which creates a fundamental obstacle to open-ended self-evolution. In this paper, we propose Tool-R0 for training general purpose tool-calling agents from scratch with self-play RL, under a zero-data assumption. Initialized from the same base LLM, Tool-R0 co-evolves a Generator and a Solver with complementary rewards: one proposes targeted challenging tasks at the other’s competence frontier and the other learns to solve them with real-world tool calls. This creates a self-evolving cycle that requires no pre-existing datasets. Evaluation on five tool-use benchmarks show that Tool-R0 yields 92.5% relative improvement over the base model and surpasses fully supervised tool-calling baselines under the same setting. Our work further provides empirical insights into self-play LLM agents by analyzing co-evolution, curriculum dynamics, and scaling behavior.</abstract>
<identifier type="citekey">acikgoz-etal-2026-tool</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1407/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>29989</start>
<end>30017</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero Data
%A Acikgoz, Emre Can
%A Qian, Cheng
%A Hübotter, Jonas
%A Ji, Heng
%A Hakkani-Tür, Dilek
%A Tur, Gokhan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F acikgoz-etal-2026-tool
%X Large language models (LLMs) are becoming the foundation for autonomous agents that can use tools to solve complex tasks. Reinforcement learning (RL) has emerged as a common approach for injecting such agentic capabilities, but typically under tightly controlled training setups. It often depends on carefully curated task–solution pairs and substantial human supervision, which creates a fundamental obstacle to open-ended self-evolution. In this paper, we propose Tool-R0 for training general purpose tool-calling agents from scratch with self-play RL, under a zero-data assumption. Initialized from the same base LLM, Tool-R0 co-evolves a Generator and a Solver with complementary rewards: one proposes targeted challenging tasks at the other’s competence frontier and the other learns to solve them with real-world tool calls. This creates a self-evolving cycle that requires no pre-existing datasets. Evaluation on five tool-use benchmarks show that Tool-R0 yields 92.5% relative improvement over the base model and surpasses fully supervised tool-calling baselines under the same setting. Our work further provides empirical insights into self-play LLM agents by analyzing co-evolution, curriculum dynamics, and scaling behavior.
%U https://aclanthology.org/2026.findings-emnlp.1407/
%P 29989-30017
Markdown (Informal)
[Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero Data](https://aclanthology.org/2026.findings-emnlp.1407/) (Acikgoz et al., Findings 2026)
ACL