大语言模型推理算法研究专家 - Seed Model
Core
Researching and improving the logical reasoning, task planning, and system-level capabilities of large language models (LLMs) to enhance their intelligence.
Role type
Senior Research Scientist (Large Language Model Reasoning)
Builds
Advanced LLM reasoning and planning capabilities integrated into products like Doubao and TRAE.
Domain
Artificial Intelligence / Large Language Models / Generative AI
Deliverable
production ML models
Required skills
Large language model research, logical reasoning, task planning, data synthesis, scalable oversight, System 2 thinking, agent/mult-agent systems, C/C++ or Python, data structures, algorithms
Preferred skills
ACM/ICPC/NOI/IOI/Top Coder/Kaggle awards, leading LLM projects or publishing papers, deep problem-solving in LLM training and application
Technologies
LLMs, System 2, Agents, Mult-Agent systems
Responsibilities
Improve LLM logic and planning across the full training pipeline (evaluation, pretraining, SFT, RLHF); Research data synthesis and scalable oversight to reduce reliance on human labeling; Apply System 2 thinking to enhance reasoning and planning; Develop robust evaluation methods for model reasoning and interaction; Enhance tool calling and API interaction capabilities via agent systems.