@inproceedings{wang-etal-2026-gts,
title = "{GTS}: Inference-Time Scaling of Latent Reasoning with a Learnable {G}aussian Thought Sampler",
author = "Wang, Minghan and
Bai, Ye and
Vu, Thuy-Trang and
Shareghi, Ehsan and
Haffari, Gholamreza",
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.341/",
pages = "7861--7884",
ISBN = "979-8-89176-459-0",
abstract = "Inference-time scaling (ITS) for latent reasoning requires sampling multiple candidate latent trajectories, yet current methods typically rely on heuristic perturbations such as dropout or fixed Gaussian noise. We show that stronger perturbations do not necessarily improve candidate-set quality: they often induce larger distribution shifts without producing more decision-useful reasoning paths. This limitation arises because heuristic perturbations inject stochasticity through fixed or hand-designed mechanisms rather than an optimizable, context-conditioned sampling policy, making latent exploration difficult to control. We focus on the sampling side of latent ITS and propose the Gaussian Thought Sampler (GTS), a lightweight module that predicts context-dependent Gaussian perturbations over continuous reasoning states. Trained with GRPO-style policy optimization while keeping the backbone frozen, GTS turns heuristic perturbation into a trainable, context-conditioned sampling policy. Experiments across two latent reasoning backbones and multiple benchmarks show that GTS consistently improves pass@N over heuristic baselines, suggesting that effective sampling for latent ITS requires controlled, optimizable exploration rather than simply amplified stochasticity."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-gts">
<titleInfo>
<title>GTS: Inference-Time Scaling of Latent Reasoning with a Learnable Gaussian Thought Sampler</title>
</titleInfo>
<name type="personal">
<namePart type="given">Minghan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ye</namePart>
<namePart type="family">Bai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Thuy-Trang</namePart>
<namePart type="family">Vu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ehsan</namePart>
<namePart type="family">Shareghi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gholamreza</namePart>
<namePart type="family">Haffari</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>Inference-time scaling (ITS) for latent reasoning requires sampling multiple candidate latent trajectories, yet current methods typically rely on heuristic perturbations such as dropout or fixed Gaussian noise. We show that stronger perturbations do not necessarily improve candidate-set quality: they often induce larger distribution shifts without producing more decision-useful reasoning paths. This limitation arises because heuristic perturbations inject stochasticity through fixed or hand-designed mechanisms rather than an optimizable, context-conditioned sampling policy, making latent exploration difficult to control. We focus on the sampling side of latent ITS and propose the Gaussian Thought Sampler (GTS), a lightweight module that predicts context-dependent Gaussian perturbations over continuous reasoning states. Trained with GRPO-style policy optimization while keeping the backbone frozen, GTS turns heuristic perturbation into a trainable, context-conditioned sampling policy. Experiments across two latent reasoning backbones and multiple benchmarks show that GTS consistently improves pass@N over heuristic baselines, suggesting that effective sampling for latent ITS requires controlled, optimizable exploration rather than simply amplified stochasticity.</abstract>
<identifier type="citekey">wang-etal-2026-gts</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.341/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>7861</start>
<end>7884</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T GTS: Inference-Time Scaling of Latent Reasoning with a Learnable Gaussian Thought Sampler
%A Wang, Minghan
%A Bai, Ye
%A Vu, Thuy-Trang
%A Shareghi, Ehsan
%A Haffari, Gholamreza
%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 wang-etal-2026-gts
%X Inference-time scaling (ITS) for latent reasoning requires sampling multiple candidate latent trajectories, yet current methods typically rely on heuristic perturbations such as dropout or fixed Gaussian noise. We show that stronger perturbations do not necessarily improve candidate-set quality: they often induce larger distribution shifts without producing more decision-useful reasoning paths. This limitation arises because heuristic perturbations inject stochasticity through fixed or hand-designed mechanisms rather than an optimizable, context-conditioned sampling policy, making latent exploration difficult to control. We focus on the sampling side of latent ITS and propose the Gaussian Thought Sampler (GTS), a lightweight module that predicts context-dependent Gaussian perturbations over continuous reasoning states. Trained with GRPO-style policy optimization while keeping the backbone frozen, GTS turns heuristic perturbation into a trainable, context-conditioned sampling policy. Experiments across two latent reasoning backbones and multiple benchmarks show that GTS consistently improves pass@N over heuristic baselines, suggesting that effective sampling for latent ITS requires controlled, optimizable exploration rather than simply amplified stochasticity.
%U https://aclanthology.org/2026.emnlp-main.341/
%P 7861-7884
Markdown (Informal)
[GTS: Inference-Time Scaling of Latent Reasoning with a Learnable Gaussian Thought Sampler](https://aclanthology.org/2026.emnlp-main.341/) (Wang et al., EMNLP 2026)
ACL