Sherpa: Teaching LLMs to Teach Adaptively

cs.AI, cs.CL

Submitted: 2026-10-06

Updated: 2026-10-06

Comments: 32 pages, 6 figures. Code and model are available at https://github.com/SALT-NLP/Sherpa

Code: https://github.com/SALT-NLP/Sherpa

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Sources

• PEARL: Training Socratic Tutors with Pedagogically Aligned Reinforcement Learning
• The Llama 3 Herd of Models
• Qwen2.5-Coder Technical Report
• Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task
• LearnLM: Improving Gemini for Learning
• Ministral 3
• GEM: A Gym for Agentic LLMs
• Eight Methods to Evaluate Robust Unlearning in LLMs
• Towards the Pedagogical Steering of Large Language Models for Tutoring: A Case Study with Modeling Productive Failure
• Simulated Students in Tutoring Dialogues: Substance or Illusion?
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Simulating Novice Students Using Machine Unlearning and Relearning in Large Language Models
• Student Data Paradox and Curious Case of Single Student-Tutor Model: Regressive Side Effects of Training LLMs for Personalized Learning
• TutorBench: A Benchmark To Assess Tutoring Capabilities Of Large Language Models
• Gemma 3 Technical Report
• Tutor CoPilot: A Human-AI Approach for Scaling Real-Time Expertise
• Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement
• Qwen3 Technical Report
• StudentSim: Training LLM-based Student Simulators

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems