Topic · 4 recaps

Alignment & Safety

Research on getting models to do what we actually want — honest, harmless, and helpful behavior — covering RLHF, constitutional methods, adversarial robustness, and broader safety theory.

Sort

Newest

StepGuard: Learning Step-Level Guardrails with Scalable Supervision and Safety-Utility Balancing

Agents · Aug 25

ComBodied Agents: a New Paradigm of Human-Centric Agentic AI

Agents · Aug 11

The Personalization Mirage: How LLMs Fabricate User Profiles, and Why Self-Monitoring Misleads

Alignment · Aug 5

BadWAM: When World-Action Models Dream Right but Act Wrong

Alignment · Jul 16

— End of list —