@inproceedings{zheng-etal-2026-progmem,
title = "{P}rog{M}em: Programmatic Belief-State Memory for Long-Horizon Conversational Agents",
author = "Zheng, Pengchao and
Wang, Xinan and
Wei, Ziyi and
Liu, Haoran and
Wei, Lei and
Niu, Jiajun and
Guo, Bohan and
Wu, Han",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {IJCNLP}-{AACL} 2026",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-aacl.120/",
pages = "2363--2376",
ISBN = "979-8-89176-444-6",
abstract = "Long-horizon conversational agents require memory that can evolve with dialogue, retrieve information on demand, and remain uncertainty-aware when evidence is insufficient, rather than merely larger storage or context windows. Existing methods store interaction histories as retrievable text, summaries, or vector indices, and retrieve relevant content at query time. Although effective for simple factual queries, this paradigm does not explicitly model how memory writes change state, leaving temporal updates, plan changes, and counterfactual reasoning to query-time inference over retrieved text.\par We propose ProgMem, which compiles dialogue events into structured memory programs and checks answers against dialogue evidence. On LoCoMo-10, ProgMem has the highest average F1 with Qwen3-4B, Qwen3-8B, and Qwen3-14B. Its temporal F1 exceeds Mem0 by up to 9.95 points and Zep by at least 12.88 points across these settings. Ablations show gains from route-aware retrieval and two-stage verification. On DialSim, ProgMem improves Answerable and FUQ F1 over Mem0 while using fewer query-time memory tokens. A single-GPU comparison finds higher mean and P95 query latency than Mem0. These results show where state maintenance helps and what it costs at query time."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zheng-etal-2026-progmem">
<titleInfo>
<title>ProgMem: Programmatic Belief-State Memory for Long-Horizon Conversational Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Pengchao</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziyi</namePart>
<namePart type="family">Wei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haoran</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lei</namePart>
<namePart type="family">Wei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Niu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bohan</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Han</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-444-6</identifier>
</relatedItem>
<abstract>Long-horizon conversational agents require memory that can evolve with dialogue, retrieve information on demand, and remain uncertainty-aware when evidence is insufficient, rather than merely larger storage or context windows. Existing methods store interaction histories as retrievable text, summaries, or vector indices, and retrieve relevant content at query time. Although effective for simple factual queries, this paradigm does not explicitly model how memory writes change state, leaving temporal updates, plan changes, and counterfactual reasoning to query-time inference over retrieved text.\par We propose ProgMem, which compiles dialogue events into structured memory programs and checks answers against dialogue evidence. On LoCoMo-10, ProgMem has the highest average F1 with Qwen3-4B, Qwen3-8B, and Qwen3-14B. Its temporal F1 exceeds Mem0 by up to 9.95 points and Zep by at least 12.88 points across these settings. Ablations show gains from route-aware retrieval and two-stage verification. On DialSim, ProgMem improves Answerable and FUQ F1 over Mem0 while using fewer query-time memory tokens. A single-GPU comparison finds higher mean and P95 query latency than Mem0. These results show where state maintenance helps and what it costs at query time.</abstract>
<identifier type="citekey">zheng-etal-2026-progmem</identifier>
<location>
<url>https://aclanthology.org/2026.findings-aacl.120/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>2363</start>
<end>2376</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ProgMem: Programmatic Belief-State Memory for Long-Horizon Conversational Agents
%A Zheng, Pengchao
%A Wang, Xinan
%A Wei, Ziyi
%A Liu, Haoran
%A Wei, Lei
%A Niu, Jiajun
%A Guo, Bohan
%A Wu, Han
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F zheng-etal-2026-progmem
%X Long-horizon conversational agents require memory that can evolve with dialogue, retrieve information on demand, and remain uncertainty-aware when evidence is insufficient, rather than merely larger storage or context windows. Existing methods store interaction histories as retrievable text, summaries, or vector indices, and retrieve relevant content at query time. Although effective for simple factual queries, this paradigm does not explicitly model how memory writes change state, leaving temporal updates, plan changes, and counterfactual reasoning to query-time inference over retrieved text.\par We propose ProgMem, which compiles dialogue events into structured memory programs and checks answers against dialogue evidence. On LoCoMo-10, ProgMem has the highest average F1 with Qwen3-4B, Qwen3-8B, and Qwen3-14B. Its temporal F1 exceeds Mem0 by up to 9.95 points and Zep by at least 12.88 points across these settings. Ablations show gains from route-aware retrieval and two-stage verification. On DialSim, ProgMem improves Answerable and FUQ F1 over Mem0 while using fewer query-time memory tokens. A single-GPU comparison finds higher mean and P95 query latency than Mem0. These results show where state maintenance helps and what it costs at query time.
%U https://aclanthology.org/2026.findings-aacl.120/
%P 2363-2376
Markdown (Informal)
[ProgMem: Programmatic Belief-State Memory for Long-Horizon Conversational Agents](https://aclanthology.org/2026.findings-aacl.120/) (Zheng et al., Findings 2026)
ACL
- Pengchao Zheng, Xinan Wang, Ziyi Wei, Haoran Liu, Lei Wei, Jiajun Niu, Bohan Guo, and Han Wu. 2026. ProgMem: Programmatic Belief-State Memory for Long-Horizon Conversational Agents. In Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026, pages 2363–2376, Hengqin, China. Association for Computational Linguistics.