@inproceedings{ding-etal-2026-mitigating,
title = "Mitigating Overthinking via Interleaved {S}ystem-1/2 Reasoning in Large Language Models",
author = "Ding, Han and
Chen, Wei and
Zhuang, Fuzhen",
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.636/",
pages = "14633--14656",
ISBN = "979-8-89176-459-0",
abstract = "Large reasoning models improve problem-solving performance by generating extended Chain-of-Thought, but often suffer from overthinking, producing unnecessarily long and redundant reasoning traces. Prior works mainly focus on directly compressing reasoning length, or selecting reasoning modes globally based on task difficulty, without accounting for the dynamic and heterogeneous nature of the reasoning process itself. This lack of consideration for inference-time diversity limits their ability to adapt reasoning strategies flexibly. To address this gap, we propose Interleaved System-1/2 Reasoning (InterSR), a cognitively inspired framework that dynamically alternates between System 1 (fast) and System 2 (slow) reasoning modes during generation. Central to InterSR is a learnable system switching signal that reflects the necessity for deeper reasoning at each intermediate step and enables fine-grained control over cognitive mode transitions. Experiments across mathematical and cross-domain benchmarks show that InterSR effectively captures internal reasoning signals to reduce token usage while maintaining strong and robust accuracy. We release our code at https://github.com/buaadh/InterSR."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ding-etal-2026-mitigating">
<titleInfo>
<title>Mitigating Overthinking via Interleaved System-1/2 Reasoning in Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Han</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fuzhen</namePart>
<namePart type="family">Zhuang</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>Large reasoning models improve problem-solving performance by generating extended Chain-of-Thought, but often suffer from overthinking, producing unnecessarily long and redundant reasoning traces. Prior works mainly focus on directly compressing reasoning length, or selecting reasoning modes globally based on task difficulty, without accounting for the dynamic and heterogeneous nature of the reasoning process itself. This lack of consideration for inference-time diversity limits their ability to adapt reasoning strategies flexibly. To address this gap, we propose Interleaved System-1/2 Reasoning (InterSR), a cognitively inspired framework that dynamically alternates between System 1 (fast) and System 2 (slow) reasoning modes during generation. Central to InterSR is a learnable system switching signal that reflects the necessity for deeper reasoning at each intermediate step and enables fine-grained control over cognitive mode transitions. Experiments across mathematical and cross-domain benchmarks show that InterSR effectively captures internal reasoning signals to reduce token usage while maintaining strong and robust accuracy. We release our code at https://github.com/buaadh/InterSR.</abstract>
<identifier type="citekey">ding-etal-2026-mitigating</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.636/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>14633</start>
<end>14656</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Mitigating Overthinking via Interleaved System-1/2 Reasoning in Large Language Models
%A Ding, Han
%A Chen, Wei
%A Zhuang, Fuzhen
%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 ding-etal-2026-mitigating
%X Large reasoning models improve problem-solving performance by generating extended Chain-of-Thought, but often suffer from overthinking, producing unnecessarily long and redundant reasoning traces. Prior works mainly focus on directly compressing reasoning length, or selecting reasoning modes globally based on task difficulty, without accounting for the dynamic and heterogeneous nature of the reasoning process itself. This lack of consideration for inference-time diversity limits their ability to adapt reasoning strategies flexibly. To address this gap, we propose Interleaved System-1/2 Reasoning (InterSR), a cognitively inspired framework that dynamically alternates between System 1 (fast) and System 2 (slow) reasoning modes during generation. Central to InterSR is a learnable system switching signal that reflects the necessity for deeper reasoning at each intermediate step and enables fine-grained control over cognitive mode transitions. Experiments across mathematical and cross-domain benchmarks show that InterSR effectively captures internal reasoning signals to reduce token usage while maintaining strong and robust accuracy. We release our code at https://github.com/buaadh/InterSR.
%U https://aclanthology.org/2026.emnlp-main.636/
%P 14633-14656
Markdown (Informal)
[Mitigating Overthinking via Interleaved System-1/2 Reasoning in Large Language Models](https://aclanthology.org/2026.emnlp-main.636/) (Ding et al., EMNLP 2026)
ACL