#dialogue

8 curated events

papersTODAY 04:00 UTC

arXiv paper proposes emotion regulation framework for empathetic speech dialogue in audio-language models

A new arXiv preprint introduces ER-EDF, a framework that draws on psychological theories of emotion perception and regulation to guide empathetic responses in spoken dialogue systems built on large audio-language models. The work aims to improve how such systems both recognize a speaker's emotional state and regulate their own generated reply. It is a research contribution and has not been presented as a product or model release.

speech/emotion/dialogue/audio-language-models

papersTODAY 04:00 UTC

Value-Guided Preference Distillation Proposed for Long-Horizon Dialogue Alignment

A new arXiv paper argues that aligning multi-turn dialogue agents by matching turn-level human preferences is a poor proxy for long-term outcomes and is vulnerable to reward hacking. The authors recast long-horizon dialogue optimization as a multi-objective problem and propose distilling dense behavioral signals into value guidance for preference-based training. The method is presented as a way to optimize sparse end goals more reliably without directly optimizing them.

alignment/dialogue/preference-optimization/research

papersTODAY 04:00 UTC

Mind2Dialogue Framework Trains Language Models to Model User Mental States

A new arXiv paper introduces Mind2Dialogue, a method that trains language models to be more aware of the people they interact with by simulating users' mental states during training. The authors frame the problem as a supervision gap: models need signals about human internal states, which are rarely available in standard dialogue data. The approach targets long-term collaboration in learning, reasoning, and decision-making tasks.

llm/dialogue/user-modeling/training

papersTODAY 04:00 UTC

ParaBridge Ties Paralinguistic Cues to Dialogue Behavior in Speech Language Models

A new arXiv paper introduces ParaBridge, a method aimed at connecting the paralinguistic information in speech — such as vocal tone, speaker traits, or background noise — with the responses a spoken dialogue system produces. The authors note that while existing speech language models can detect such cues, they often fail to let that perception shape their replies, and the work targets closing that gap.

speech/dialogue/research/multimodal

papersTODAY 04:00 UTC

Paper Proposes Self-Play Method for Training AI Assistants When to Ask Clarifying Questions

A new arXiv preprint introduces a method that teaches AI assistants how to handle ambiguous or underspecified user requests. The approach uses collaborative self-play to learn a steerable policy that decides between answering directly, listing several possible interpretations, or asking the user for clarification. The goal is to improve how systems manage uncertainty in dialogue rather than guessing wrong.

research/agents/self-play/dialogue

papersSEP 10 04:00 UTC

SocialRL trains LLMs' social intelligence with multi-turn reinforcement learning and reward design

A new arXiv paper presents SocialRL, a framework that applies multi-turn reinforcement learning together with carefully designed rewards to sharpen how language models handle social context in extended conversations. The work targets agents' capacity to read situational cues, infer speaker intent, and adjust behavior over sustained dialogue, with the goal of more effective and trustworthy human-AI collaboration.

reinforcement-learning/agents/dialogue/human-ai-interaction

papersSEP 10 04:00 UTC

RelayS2S: Dual-Path Speculative Generation for Real-Time Speech-to-Speech Dialogue

A new arXiv paper proposes RelayS2S, a dual-path speculative generation method for real-time spoken dialogue systems. It addresses the trade-off between latency and response quality, since end-to-end speech-to-speech models can respond instantly and manage turn-taking, backchanneling, and interruptions, but tend to produce semantically weaker replies. The method aims to combine immediate responsiveness with improved response content.

speech-to-speech/real-time/dialogue/latency

papersSEP 10 04:00 UTC

arXiv Study Presents Expert-Level Crisis Detection in Mental Health Conversations

A newly updated arXiv paper tackles the problem of spotting mental health crisis situations during live, multi-turn conversations instead of isolated snippets of text. The authors note that current models lose considerable accuracy when risk indicators unfold across dialogue turns, and they introduce an approach aimed at expert-level detection in these conversational settings.

mental-health/safety/nlp/dialogue