@inproceedings{gao-etal-2026-proceedrl,
title = "{P}ro{C}eed{RL}: Process Critic with Explorative Demonstration Reinforcement Learning for {LLM} Agentic Reasoning",
author = "Gao, Jingyue and
Guo, Yanjiang and
Xiaoshuai, Chen and
Chen, Jianyu",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.800/",
doi = "10.18653/v1/2026.findings-acl.800",
pages = "16281--16295",
ISBN = "979-8-89176-395-1",
abstract = "Reinforcement Learning (RL) significantly enhances the reasoning abilities of large language models (LLMs), yet applying it to multi-turn agentic tasks remains challenging due to the long-horizon nature of interactions and the stochasticity of environmental feedback.We identify a structural failure mode in agentic exploration: suboptimal actions elicit noisy observations into misleading contexts, which further weaken subsequent decision-making, making recovery increasingly difficult.This cumulative feedback loop of errors renders standard exploration strategies ineffective and susceptible to the model{'}s reasoning and the environment{'}s randomness.To mitigate this issue, we propose \textbf{ProCeedRL}: \textbf{Pro}cess \textbf{C}ritic with \textbf{E}xplorativ\textbf{e} \textbf{D}emonstration RL, shifting exploration from passive selection to active intervention.ProCeedRL employs a process-level critic to monitor interactions in real time, incorporating reflection-based demonstrations to guide agents in stopping the accumulation of errors.We find that this approach significantly exceeds the model{'}s saturated exploration performance, demonstrating substantial exploratory benefits.By learning from exploratory demonstrations and on-policy samples, ProCeedRL significantly improves exploration efficiency and achieves superior performance on complex deep search and embodied tasks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="gao-etal-2026-proceedrl">
<titleInfo>
<title>ProCeedRL: Process Critic with Explorative Demonstration Reinforcement Learning for LLM Agentic Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jingyue</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanjiang</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chen</namePart>
<namePart type="family">Xiaoshuai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jianyu</namePart>
<namePart type="family">Chen</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>Findings of the Association for Computational Linguistics: ACL 2026</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">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Reinforcement Learning (RL) significantly enhances the reasoning abilities of large language models (LLMs), yet applying it to multi-turn agentic tasks remains challenging due to the long-horizon nature of interactions and the stochasticity of environmental feedback.We identify a structural failure mode in agentic exploration: suboptimal actions elicit noisy observations into misleading contexts, which further weaken subsequent decision-making, making recovery increasingly difficult.This cumulative feedback loop of errors renders standard exploration strategies ineffective and susceptible to the model’s reasoning and the environment’s randomness.To mitigate this issue, we propose ProCeedRL: Process Critic with Explorative Demonstration RL, shifting exploration from passive selection to active intervention.ProCeedRL employs a process-level critic to monitor interactions in real time, incorporating reflection-based demonstrations to guide agents in stopping the accumulation of errors.We find that this approach significantly exceeds the model’s saturated exploration performance, demonstrating substantial exploratory benefits.By learning from exploratory demonstrations and on-policy samples, ProCeedRL significantly improves exploration efficiency and achieves superior performance on complex deep search and embodied tasks.</abstract>
<identifier type="citekey">gao-etal-2026-proceedrl</identifier>
<identifier type="doi">10.18653/v1/2026.findings-acl.800</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.800/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>16281</start>
<end>16295</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ProCeedRL: Process Critic with Explorative Demonstration Reinforcement Learning for LLM Agentic Reasoning
%A Gao, Jingyue
%A Guo, Yanjiang
%A Xiaoshuai, Chen
%A Chen, Jianyu
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F gao-etal-2026-proceedrl
%X Reinforcement Learning (RL) significantly enhances the reasoning abilities of large language models (LLMs), yet applying it to multi-turn agentic tasks remains challenging due to the long-horizon nature of interactions and the stochasticity of environmental feedback.We identify a structural failure mode in agentic exploration: suboptimal actions elicit noisy observations into misleading contexts, which further weaken subsequent decision-making, making recovery increasingly difficult.This cumulative feedback loop of errors renders standard exploration strategies ineffective and susceptible to the model’s reasoning and the environment’s randomness.To mitigate this issue, we propose ProCeedRL: Process Critic with Explorative Demonstration RL, shifting exploration from passive selection to active intervention.ProCeedRL employs a process-level critic to monitor interactions in real time, incorporating reflection-based demonstrations to guide agents in stopping the accumulation of errors.We find that this approach significantly exceeds the model’s saturated exploration performance, demonstrating substantial exploratory benefits.By learning from exploratory demonstrations and on-policy samples, ProCeedRL significantly improves exploration efficiency and achieves superior performance on complex deep search and embodied tasks.
%R 10.18653/v1/2026.findings-acl.800
%U https://aclanthology.org/2026.findings-acl.800/
%U https://doi.org/10.18653/v1/2026.findings-acl.800
%P 16281-16295
Markdown (Informal)
[ProCeedRL: Process Critic with Explorative Demonstration Reinforcement Learning for LLM Agentic Reasoning](https://aclanthology.org/2026.findings-acl.800/) (Gao et al., Findings 2026)
ACL