@inproceedings{guo-etal-2026-answer,
title = "The Answer Lies Within: Self-Derived Rewards Enable Explainable Relation Extraction",
author = "Guo, Xinyu and
Shi, Zhengliang and
Yang, Minglai and
Rahimi, Mahdi and
Surdeanu, Mihai",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2284/",
pages = "51957--51993",
ISBN = "979-8-89176-459-0",
abstract = "Despite the remarkable reasoning capabilities of large language models, they still struggle with one-shot relation extraction without predefined relation labels. We identify two pitfalls: models are often misled by irrelevant tokens instead of relation-conveying semantics, and they often fail to align with the abstraction level human annotators expect. We introduce a novel framework that closes this gap with two components: (1) CogRE, a cognitively-inspired reasoning framework that structures RE into a series of processes mimicking human text-processing; and (2) Hit@Dict, a reinforcement learning intermediate reward strategy that ncourages reasoning to align with relational labels by rewarding relation-relevant phrases in reasoning. The reward is based on a credit dictionary automatically extracted from gold outputs. Our experiments show that our framework improves both accuracy and explanation quality by addressing these two pitfalls. For example, CogRE with Qwen2.5-14B-Instruct on One-shot NYT29 achieves 24.65{\%} F1, surpassing prior reasoning-based designs. Optimizing this approach with RL using Hit@Dict further improves performance by +23.46{\%} points. Finally, human evaluation shows that our best model generates relational phrases closely aligned with gold labels, increasing human explanation quality ratings by 54{\%} (relative)."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="guo-etal-2026-answer">
<titleInfo>
<title>The Answer Lies Within: Self-Derived Rewards Enable Explainable Relation Extraction</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xinyu</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhengliang</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Minglai</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mahdi</namePart>
<namePart type="family">Rahimi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mihai</namePart>
<namePart type="family">Surdeanu</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Despite the remarkable reasoning capabilities of large language models, they still struggle with one-shot relation extraction without predefined relation labels. We identify two pitfalls: models are often misled by irrelevant tokens instead of relation-conveying semantics, and they often fail to align with the abstraction level human annotators expect. We introduce a novel framework that closes this gap with two components: (1) CogRE, a cognitively-inspired reasoning framework that structures RE into a series of processes mimicking human text-processing; and (2) Hit@Dict, a reinforcement learning intermediate reward strategy that ncourages reasoning to align with relational labels by rewarding relation-relevant phrases in reasoning. The reward is based on a credit dictionary automatically extracted from gold outputs. Our experiments show that our framework improves both accuracy and explanation quality by addressing these two pitfalls. For example, CogRE with Qwen2.5-14B-Instruct on One-shot NYT29 achieves 24.65% F1, surpassing prior reasoning-based designs. Optimizing this approach with RL using Hit@Dict further improves performance by +23.46% points. Finally, human evaluation shows that our best model generates relational phrases closely aligned with gold labels, increasing human explanation quality ratings by 54% (relative).</abstract>
<identifier type="citekey">guo-etal-2026-answer</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2284/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>51957</start>
<end>51993</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The Answer Lies Within: Self-Derived Rewards Enable Explainable Relation Extraction
%A Guo, Xinyu
%A Shi, Zhengliang
%A Yang, Minglai
%A Rahimi, Mahdi
%A Surdeanu, Mihai
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F guo-etal-2026-answer
%X Despite the remarkable reasoning capabilities of large language models, they still struggle with one-shot relation extraction without predefined relation labels. We identify two pitfalls: models are often misled by irrelevant tokens instead of relation-conveying semantics, and they often fail to align with the abstraction level human annotators expect. We introduce a novel framework that closes this gap with two components: (1) CogRE, a cognitively-inspired reasoning framework that structures RE into a series of processes mimicking human text-processing; and (2) Hit@Dict, a reinforcement learning intermediate reward strategy that ncourages reasoning to align with relational labels by rewarding relation-relevant phrases in reasoning. The reward is based on a credit dictionary automatically extracted from gold outputs. Our experiments show that our framework improves both accuracy and explanation quality by addressing these two pitfalls. For example, CogRE with Qwen2.5-14B-Instruct on One-shot NYT29 achieves 24.65% F1, surpassing prior reasoning-based designs. Optimizing this approach with RL using Hit@Dict further improves performance by +23.46% points. Finally, human evaluation shows that our best model generates relational phrases closely aligned with gold labels, increasing human explanation quality ratings by 54% (relative).
%U https://aclanthology.org/2026.emnlp-main.2284/
%P 51957-51993
Markdown (Informal)
[The Answer Lies Within: Self-Derived Rewards Enable Explainable Relation Extraction](https://aclanthology.org/2026.emnlp-main.2284/) (Guo et al., EMNLP 2026)
ACL