OnePred: Next-Query Prediction via Recursive Intent Memory in Multi-Turn Conversations

• URL: http://arxiv.org/abs/2605.23668v1
• Date: Fri, 22 May 2026 14:16:21 GMT
• Title: OnePred: Next-Query Prediction via Recursive Intent Memory in Multi-Turn Conversations
• Abstract summary: Large language model (LLM) conversational systems process millions of multi-turn dialogues daily.<n>Next-Query prediction anticipates the user's subsequent query based solely on the preceding dialogue.<n>OnePred tracks the user's evolving intent trajectory across topics, unresolved needs, and interest shifts.
• Score: 6.89645931986174
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Although large language model (LLM) conversational systems process millions of multi-turn dialogues daily, they remain fundamentally reactive: they respond only after the user types a query. A key step toward proactive interaction is next-query prediction, which anticipates the user's subsequent query based solely on the preceding dialogue. Progress on this task is hindered by the lack of dedicated benchmarks and a fundamental efficiency--quality trade-off: naively concatenating full dialogue history incurs linearly growing token consumption, while truncating to the latest turn discards crucial cross-turn context. Our key insight is that accurate prediction does not require re-reading raw history; it suffices to track the user's evolving intent trajectory across topics, unresolved needs, and interest shifts. We propose OnePred, which maintains a recursively updated memory as its sole cross-turn context, bounding the per-turn cost independently of conversation length. We train the model via a two-stage reinforcement learning pipeline that first teaches what to predict, then what to compress, shaping the memory into a prediction-oriented intent chain. To establish a rigorous testbed, we introduce NQP-Bench, spanning three diverse subsets. Experiments demonstrate that OnePred reduces per-turn token consumption by up to 22$\times$ compared to full-history inputs while consistently exceeding all baselines in prediction quality, with larger gains on longer conversations. Our code is publicly available at https://github.com/ZBWpro/OnePred.

Related papers

• PERMA: Benchmarking Personalized Memory Agents via Event-Driven Preference and Realistic Task Environments [72.02445514666428]
  We introduce PERMA, a benchmark designed to evaluate persona consistency over time beyond static preference recall.<n>PerMA consists of temporally ordered interaction events spanning multiple sessions and domains, with preference-related queries inserted over time.<n> Experiments demonstrate that by linking related interactions, advanced memory systems can extract more precise preferences and reduce token consumption.
  arXiv  Detail & Related papers  (2026-03-24T14:04:11Z)
• Do LLMs Benefit From Their Own Words? [56.73014497206615]
  We find that removing prior assistant responses does not affect response quality on a large fraction of turns.<n>Omitting assistant-side context can reduce cumulative context lengths by up to 10x.<n>Our findings suggest that selectively omitting assistant history can improve response quality while reducing memory consumption.
  arXiv  Detail & Related papers  (2026-02-27T18:58:26Z)
• LoopServe: An Adaptive Dual-phase LLM Inference Acceleration System for Multi-Turn Dialogues [59.12542274007847]
  LoopServe is an adaptive dual-phase inference framework for large language models in multi-turn dialogues.<n>LoopServe consistently achieves superior effectiveness compared to existing baselines.
  arXiv  Detail & Related papers  (2025-07-18T06:12:08Z)
• Phrase Retrieval for Open-Domain Conversational Question Answering with Conversational Dependency Modeling via Contrastive Learning [54.55643652781891]
  Open-Domain Conversational Question Answering (ODConvQA) aims at answering questions through a multi-turn conversation. We propose a method to directly predict answers with a phrase retrieval scheme for a sequence of words.
  arXiv  Detail & Related papers  (2023-06-07T09:46:38Z)
• Re$^3$Dial: Retrieve, Reorganize and Rescale Dialogue Corpus for Long-Turn Open-Domain Dialogue Pre-training [90.3412708846419]
  Most dialogues in existing pre-training corpora contain fewer than three turns of dialogue. We propose the Retrieve, Reorganize and Rescale framework (Re$3$Dial) to automatically construct billion-scale long-turn dialogues. By repeating the above process, Re$3$Dial can yield a coherent long-turn dialogue.
  arXiv  Detail & Related papers  (2023-05-04T07:28:23Z)
• A Graph-Based Context-Aware Model to Understand Online Conversations [3.8345539498627437]
  In online conversations, comments and replies may be based on external context beyond the immediately relevant information. We propose GraphNLI, a novel graph-based deep learning architecture that uses graph walks to incorporate the wider context of a conversation. We evaluate GraphNLI on two such tasks - polarity prediction and misogynistic hate speech detection.
  arXiv  Detail & Related papers  (2022-11-16T20:51:45Z)
• Sparse and Dense Approaches for the Full-rank Retrieval of Responses for Dialogues [11.726528038065764]
  We focus on the more realistic task of full-rank retrieval of responses, where $n$ can be up to millions of responses. Our findings based on three different information-seeking dialogue datasets reveal that a learned response expansion technique is a solid baseline for sparse retrieval. We find the best performing method overall to be dense retrieval with intermediate training, followed by fine-tuning on the target conversational data.
  arXiv  Detail & Related papers  (2022-04-22T08:15:15Z)
• Progressive End-to-End Object Detection in Crowded Scenes [96.92416613336096]
  Previous query-based detectors suffer from two drawbacks: first, multiple predictions will be inferred for a single object, typically in crowded scenes; second, the performance saturates as the depth of the decoding stage increases. We propose a progressive predicting method to address the above issues. Specifically, we first select accepted queries to generate true positive predictions, then refine the rest noisy queries according to the previously accepted predictions. Experiments show that our method can significantly boost the performance of query-based detectors in crowded scenes.
  arXiv  Detail & Related papers  (2022-03-15T06:12:00Z)
• Improved Goal Oriented Dialogue via Utterance Generation and Look Ahead [5.062869359266078]
  intent prediction can be improved by training a deep text-to-text neural model to generate successive user utterances from unlabeled dialogue data. We present a novel look-ahead approach that uses user utterance generation to improve intent prediction in time.
  arXiv  Detail & Related papers  (2021-10-24T11:12:48Z)
• CloneBot: Personalized Dialogue-Response Predictions [0.0]
  The project task was to create a model that, given a speaker ID, chat history, and an utterance query, can predict the response utterance in a conversation. The model is personalized for each speaker. This task can be a useful tool for building speech bots that talk in a human-like manner in a live conversation.
  arXiv  Detail & Related papers  (2021-03-31T01:15:37Z)
• Dialogue History Matters! Personalized Response Selectionin Multi-turn Retrieval-based Chatbots [62.295373408415365]
  We propose a personalized hybrid matching network (PHMN) for context-response matching. Our contributions are two-fold: 1) our model extracts personalized wording behaviors from user-specific dialogue history as extra matching information. We evaluate our model on two large datasets with user identification, i.e., personalized dialogue Corpus Ubuntu (P- Ubuntu) and personalized Weibo dataset (P-Weibo)
  arXiv  Detail & Related papers  (2021-03-17T09:42:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.