Research map / Large language models / AI and society
An LLM-Guided Tutoring System for Social Skills Training
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Introduces GLOSS, an instructor-in-the-loop LLM tutoring framework that co-builds branching narrative graphs for social skills training, with GPT-4-powered simulation, feedback, and visualization.
Problem
Traditional classroom social skills training lacks realistic, one-to-one interactive scenarios, and existing intelligent tutoring systems are hard for instructors to adapt without programming skills.
Approach
Instructors and an LLM collaboratively build a narrative graph of scenario branches; the LLM powers an avatar dialogue simulator, generates new branches in real time when student responses don't match existing options, gives feedback via an independent prompt, and an analysis tool visualizes the student's path through the graph.
Key results
- No quantitative results are reported; the paper describes the system design and states user studies are future work.
Limitations
- No user study or empirical evaluation of effectiveness reported
- Avatar responses need enhancement and balance between scripted and open interactions is unresolved
- Feedback mechanisms (immediate and delayed) still to be refined
Experimental setup
Builds on: GENEVA
Closest papers by meaning
- SocraticLM: Exploring Socratic Personalized Teaching with Large Language Models (NeurIPS 2024 · similarity 0.84)
- Can LLMs Reliably Simulate Human Learner Actions? A Simulation Authoring Framework for Open-Ended Learning Environments (AAAI 2025 · similarity 0.83)
- DICE: Decoupling Capability from Intervention Necessity in LLM Tutoring (NeurIPS 2026 · similarity 0.83)
- Teaching LLMs How to Learn with Contextual Fine-Tuning (ICLR 2025 · similarity 0.83)
- BEAGLE: Behavior-Enforced Agent for Grounded Learner Emulation (NeurIPS 2026 · similarity 0.82)
- Strong Teacher Not Needed? On Distillation in LLM Pretraining (NeurIPS 2026 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.