LLM Agent Research Intern
Core
Train LLM agents to play in-game characters with consistent personas, memory, and world rules using supervised fine-tuning and reinforcement learning.
Role type
Research intern (LLM agents & role-play)
Builds
Live role-play NPC models for games
Domain
Gaming + Large Language Models
Deliverable
production ML models
Required skills
supervised fine-tuning (SFT), direct preference optimization (DPO), reinforcement learning with verifiable rewards (RLVR), Python, PyTorch, Hugging Face/Transformers, persona consistency, multi-turn dialogue
Preferred skills
LLM agents (memory, planning, tool use), LLM applications in games, publications at major NLP/ML venues
Technologies
Python, PyTorch, Hugging Face, Transformers
Responsibilities
Train role-play agents with SFT, DPO, and RLVR; Build and curate SFT, preference, and verifiable-reward data; Design automatic checks for format, language, persona boundaries, and safety; Collaborate with game designers and engineers to integrate agents into live games
Seniority
Intern
