@inproceedings{zhen-etal-2026-hierarchical,
title = "Hierarchical Reinforcement Learning with Augmented Step-Level Transitions for {LLM} Agents",
author = "Zhen, Shuai and
Yu, Yanhua and
Guo, Ruopei and
Cheng, Nan and
Deng, Yang",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.318/",
pages = "7035--7053",
ISBN = "979-8-89176-390-6",
abstract = "Large language model (LLM) agents have demonstrated strong capabilities in complex interactive decision-making tasks.However, existing LLM agents typically rely on increasingly long interaction histories, resulting in high computational cost and limited scalability.In this paper, we propose **STEP-HRL**, a hierarchical reinforcement learning (HRL) framework that enables step-level learning by conditioning only on single-step transitions rather than full interaction histories.STEP-HRL structures tasks hierarchically, using completed subtasks to represent *global progress* of overall task. By introducing a *local progress* module, it also iteratively and selectively summarizes interaction history within each subtask to produce a compact summary of local progress.Together, these components yield augmented step-level transitions for both high-level and low-level policies.Experimental results on ScienceWorld and ALFWorld benchmarks consistently demonstrate that STEP-HRL substantially outperforms baselines in terms of performance and generalization while reducing token usage. Our code is available at https://github.com/TonyStark042/STEP-HRL."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhen-etal-2026-hierarchical">
<titleInfo>
<title>Hierarchical Reinforcement Learning with Augmented Step-Level Transitions for LLM Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shuai</namePart>
<namePart type="family">Zhen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanhua</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruopei</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nan</namePart>
<namePart type="family">Cheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Deng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Large language model (LLM) agents have demonstrated strong capabilities in complex interactive decision-making tasks.However, existing LLM agents typically rely on increasingly long interaction histories, resulting in high computational cost and limited scalability.In this paper, we propose **STEP-HRL**, a hierarchical reinforcement learning (HRL) framework that enables step-level learning by conditioning only on single-step transitions rather than full interaction histories.STEP-HRL structures tasks hierarchically, using completed subtasks to represent *global progress* of overall task. By introducing a *local progress* module, it also iteratively and selectively summarizes interaction history within each subtask to produce a compact summary of local progress.Together, these components yield augmented step-level transitions for both high-level and low-level policies.Experimental results on ScienceWorld and ALFWorld benchmarks consistently demonstrate that STEP-HRL substantially outperforms baselines in terms of performance and generalization while reducing token usage. Our code is available at https://github.com/TonyStark042/STEP-HRL.</abstract>
<identifier type="citekey">zhen-etal-2026-hierarchical</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.318/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>7035</start>
<end>7053</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Hierarchical Reinforcement Learning with Augmented Step-Level Transitions for LLM Agents
%A Zhen, Shuai
%A Yu, Yanhua
%A Guo, Ruopei
%A Cheng, Nan
%A Deng, Yang
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F zhen-etal-2026-hierarchical
%X Large language model (LLM) agents have demonstrated strong capabilities in complex interactive decision-making tasks.However, existing LLM agents typically rely on increasingly long interaction histories, resulting in high computational cost and limited scalability.In this paper, we propose **STEP-HRL**, a hierarchical reinforcement learning (HRL) framework that enables step-level learning by conditioning only on single-step transitions rather than full interaction histories.STEP-HRL structures tasks hierarchically, using completed subtasks to represent *global progress* of overall task. By introducing a *local progress* module, it also iteratively and selectively summarizes interaction history within each subtask to produce a compact summary of local progress.Together, these components yield augmented step-level transitions for both high-level and low-level policies.Experimental results on ScienceWorld and ALFWorld benchmarks consistently demonstrate that STEP-HRL substantially outperforms baselines in terms of performance and generalization while reducing token usage. Our code is available at https://github.com/TonyStark042/STEP-HRL.
%U https://aclanthology.org/2026.acl-long.318/
%P 7035-7053
Markdown (Informal)
[Hierarchical Reinforcement Learning with Augmented Step-Level Transitions for LLM Agents](https://aclanthology.org/2026.acl-long.318/) (Zhen et al., ACL 2026)
ACL