DiscussLLM introduces a two-stage synthetic data pipeline to annotate multi-turn discussions with five intervention types and trains LLMs to time contributions via a silent token or proactive responses.
arXiv:2405.13003 [cs.CL]
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2025 2verdicts
UNVERDICTED 2representative citing papers
This survey synthesizes user simulation across AI, HCI, IR, and related fields, framing a shift to generative approaches, ethical uses, AGI connections, and an academic-industry ecosystem.
citing papers explorer
-
DiscussLLM: Teaching Large Language Models When to Speak
DiscussLLM introduces a two-stage synthetic data pipeline to annotate multi-turn discussions with five intervention types and trains LLMs to time contributions via a silent token or proactive responses.
-
User Simulation in the Era of Generative AI: User Modeling, Synthetic Data Generation, and System Evaluation
This survey synthesizes user simulation across AI, HCI, IR, and related fields, framing a shift to generative approaches, ethical uses, AGI connections, and an academic-industry ecosystem.