Helps AI find the right pages in long documents by giving it richer clues than raw text alone.

Teaches voice AI to hear a patient’s cough as a clinical signal instead of background noise.

A unified framework for understanding on-policy self-distillation as dense teacher-guided supervision in RL-based LLM post-training.

Tests whether AI agents can actually do real, paid professional work — not just pass toy benchmarks.

A simple inference add-on that lets AI models handle long conversations with far less memory.

Patients and eye doctors gave high marks to a voice AI that teaches patients about eye-injection treatment.

A way to install a new behavior in an AI model — and cleanly switch it off later.

Makes the newest class of large AI models run significantly faster on the same hardware.

Helps AI answer complex questions by pulling knowledge from many angles at once instead of one.

Teaches AI models to keep the right memories — and drop the rest — during long conversations.