Mission: Exploring innovative approaches to advance AI research. Dedicated to pushing boundaries in machine learning. Committed to developing robust and ethical AI solutions. Striving to make meaningful contributions to the field.

Mission: I aim to ensure AI systems are safe, aligned, and beneficial by developing principled alignment mechanisms and exploring large model applications in socially impactful domains such as healthcare, education, and science.

Mission: Learning from human feedback is the key to the continuous progress of AI. I am dedicated to providing richer feedback for AI, such as natural language and formal languages, to empower AI alignment and AI safety.

Mission: My research focuses on the safety and capabilities of AI systems. This includes the accurate and scalable evaluation of AI systems, as well as ensuring they align with human intent and values through mechanism design.

Mission: Exploring innovative approaches to advance AI research. Dedicated to pushing boundaries in machine learning. Committed to developing robust and ethical AI solutions. Striving to make meaningful contributions to the field.

Mission: Develop scalable oversight and moral alignment mechanisms that integrate theoretical and empirical approaches to ensure ethically grounded, socially responsible intelligence beyond human-level capabilities.

Mission: Exploring innovative approaches to advance AI research. Dedicated to pushing boundaries in machine learning. Committed to developing robust and ethical AI solutions. Striving to make meaningful contributions to the field.

Mission: To facilitate human moral progress with truth-seeking AI. Pervasive AI influence is harming the epistemics of the human-AI collective, and I hope to reverse the trend and turn them into facilitators of collective reflection.

Mission: Supervision signals are fundamental to AI alignment and safety. My research aims to develop scalable, stable, and verifiable supervision through formal verification and mechanistic interpretability, and to leverage reinforcement learning for optimizing the utilization of these signals in frontier AI systems.

Mission: To develop principled and provably safe foundations for intelligent systems, to foster transparent, reliable, and value-aligned AI that can be trusted in high-stakes real-world environments.

Mission: Turning scientific vision into engineered reality.

Mission: Devoted to finding and overseeing advanced AI failure modes, aiming to proactively identify and mitigate risks to ensure the safe and reliable development of artificial intelligence.