@inproceedings{peng-etal-2026-incentivizing,
title = "Incentivizing In-depth Reasoning over Long Contexts with Process Advantage Shaping",
author = "Peng, Miao and
Shen, Weizhou and
Chen, Nuo and
Li, Chenliang and
Yan, Ming and
Li, Jia",
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.1306/",
pages = "26201--26228",
ISBN = "979-8-89176-395-1",
abstract = "Reinforcement Learning with Verifiable Rewards (RLVR) has proven effective in enhancing LLMs' short-context reasoning but falters in long-context scenarios requiring precise grounding and multi-hop reasoning. We identify the ``almost-there'' phenomenon{---}trajectories that are largely correct but fail at the final step{---}in long-context reasoning RL and attribute this failure to two factors: (1) the lack of high reasoning density in long-context QA data, and (2) indiscriminate penalization of partially correct trajectories during long-context RL. To overcome this bottleneck, we propose DeepReasonQA, a KG-driven synthesis framework that controllably constructs high-difficulty, multi-hop long-context QA pairs with inherent reasoning chains. Building on this, we introduce Long-context Process Advantage Shaping (LongPAS), a simple yet effective method that performs fine-grained credit assignment by measuring reasoning steps along Validity and Relevance dimensions, which captures critical signals from ``almost-there'' trajectories. Experiments on three long-context reasoning benchmarks show that our approach substantially outperforms RLVR baselines and matches frontier LLMs while using far fewer parameters. Further analysis confirms the effectiveness of our methods in strengthening long-context reasoning while maintaining stable RL training."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="peng-etal-2026-incentivizing">
<titleInfo>
<title>Incentivizing In-depth Reasoning over Long Contexts with Process Advantage Shaping</title>
</titleInfo>
<name type="personal">
<namePart type="given">Miao</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weizhou</namePart>
<namePart type="family">Shen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nuo</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chenliang</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ming</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jia</namePart>
<namePart type="family">Li</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 with Verifiable Rewards (RLVR) has proven effective in enhancing LLMs’ short-context reasoning but falters in long-context scenarios requiring precise grounding and multi-hop reasoning. We identify the “almost-there” phenomenon—trajectories that are largely correct but fail at the final step—in long-context reasoning RL and attribute this failure to two factors: (1) the lack of high reasoning density in long-context QA data, and (2) indiscriminate penalization of partially correct trajectories during long-context RL. To overcome this bottleneck, we propose DeepReasonQA, a KG-driven synthesis framework that controllably constructs high-difficulty, multi-hop long-context QA pairs with inherent reasoning chains. Building on this, we introduce Long-context Process Advantage Shaping (LongPAS), a simple yet effective method that performs fine-grained credit assignment by measuring reasoning steps along Validity and Relevance dimensions, which captures critical signals from “almost-there” trajectories. Experiments on three long-context reasoning benchmarks show that our approach substantially outperforms RLVR baselines and matches frontier LLMs while using far fewer parameters. Further analysis confirms the effectiveness of our methods in strengthening long-context reasoning while maintaining stable RL training.</abstract>
<identifier type="citekey">peng-etal-2026-incentivizing</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.1306/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>26201</start>
<end>26228</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Incentivizing In-depth Reasoning over Long Contexts with Process Advantage Shaping
%A Peng, Miao
%A Shen, Weizhou
%A Chen, Nuo
%A Li, Chenliang
%A Yan, Ming
%A Li, Jia
%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 peng-etal-2026-incentivizing
%X Reinforcement Learning with Verifiable Rewards (RLVR) has proven effective in enhancing LLMs’ short-context reasoning but falters in long-context scenarios requiring precise grounding and multi-hop reasoning. We identify the “almost-there” phenomenon—trajectories that are largely correct but fail at the final step—in long-context reasoning RL and attribute this failure to two factors: (1) the lack of high reasoning density in long-context QA data, and (2) indiscriminate penalization of partially correct trajectories during long-context RL. To overcome this bottleneck, we propose DeepReasonQA, a KG-driven synthesis framework that controllably constructs high-difficulty, multi-hop long-context QA pairs with inherent reasoning chains. Building on this, we introduce Long-context Process Advantage Shaping (LongPAS), a simple yet effective method that performs fine-grained credit assignment by measuring reasoning steps along Validity and Relevance dimensions, which captures critical signals from “almost-there” trajectories. Experiments on three long-context reasoning benchmarks show that our approach substantially outperforms RLVR baselines and matches frontier LLMs while using far fewer parameters. Further analysis confirms the effectiveness of our methods in strengthening long-context reasoning while maintaining stable RL training.
%U https://aclanthology.org/2026.findings-acl.1306/
%P 26201-26228
Markdown (Informal)
[Incentivizing In-depth Reasoning over Long Contexts with Process Advantage Shaping](https://aclanthology.org/2026.findings-acl.1306/) (Peng et al., Findings 2026)
ACL