aiqbench analysis
This research proposes Latent-IM, a framework designed to recover explicit dialogue control mechanisms within the hidden representations of large language models for speech tasks. By decoupling move selection from generation, the method improves conversational accuracy without requiring full fine-tuning, offering new insights into steering LLM behavior during interaction.
Latent-IM: Latent Interaction Management for Speech LLMs
arXiv:2607.26928v1 Announce Type: new Abstract: Classical spoken dialogue systems often separated dialogue management from response realization: a policy selected the next dialogue action, and a generation component expressed that action. As dialogue systems shift toward LLMs, this decomposition has largely disappeared into the model's hidden representations. We ask whether an LLM-internal analogue of state estimation and action control can be recovered for conversational moves such as acknowledging, checking, querying, explaining, and replying. We formulate move control as two coupled problems: selection, predicting the appropriate next move from the dialogue context, and realization, causally producing a chosen move at generation time. We introduce Latent-IM, an internal dialogue-management framework that provides a general interface for choosing and deploying conversational moves under different objectives. Here, we use this control to reproduce human move choices, improving average end-to-end move accuracy by 12.5 points over the unsteered backbone while performing comparably to fine-tuning.
Abstract:Classical spoken dialogue systems often separated dialogue management from response realization: a policy selected the next dialogue action, and a generation component expressed that action. As dialogue systems shift toward LLMs, this decomposition has largely disappeared into the model's hidden representations. We ask whether an LLM-internal analogue of state estimation and action control can be recovered for conversational moves such as acknowledging, checking, querying, explaining, and replying. We formulate move control as two coupled problems: selection, predicting the appropriate next move from the dialogue context, and realization, causally producing a chosen move at generation time. We introduce Latent-IM, an internal dialogue-management framework that provides a general interface for choosing and deploying conversational moves under different objectives. Here, we use this control to reproduce human move choices, improving average end-to-end move accuracy by 12.5 points over the unsteered backbone while performing comparably to fine-tuning.
Computation and Language (cs.CL)
arXiv:2607.26928 [cs.CL]
(or arXiv:2607.26928v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2607.26928 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 238 words
Related coverage
[AINews] The Future of Latent Space
Latent Space · Sep 25, 2026
Pathway’s brain-inspired architecture development on Amazon SageMaker HyperPod
AWS · Sep 8, 2026
[AINews] Opus 5.5 is good at explainer videos
Latent Space · Sep 29, 2026
LLM-Based Synthetic Ground Truth Generation for Audio-Based Emotion Classification via In-Context Learning
arXiv · Jun 16, 2026