Microsoft researchers have developed On-Policy Context Distillation (OPCD), a training method that permanently embeds ...

Andrew Ting MD is part of a new generation of clinician-researchers—doctors who respect the human side of therapy while ...

Microsoft's Phi-4-reasoning-vision-15B uses careful data curation and selective reasoning to compete with models trained on ...

March 9, 2026) - Camp Lucky Board and Train announced the successful completion of its recent three-week residential board and train program conducted in Greensboro, North Carolina. The program ...

MIT researchers developed Attention Matching, a KV cache compaction technique that compresses LLM memory by 50x in seconds — ...

Facing strict privacy laws, telcos use AI-generated synthetic data as a compliant workaround to train ML models without exposing sensitive customer information.

New research was motivated by "high-potential societal impacts" of successfully disabling autonomous drones, a study author ...

Anthropic identifies AI persona drift and ties it to an “assistant axis”; tests across 275 roleplay characters, raising safety limits.

According to the researchers, the ultimate goal is to build a comprehensive cyber threat intelligence ecosystem for artificial intelligence systems. Such a system would allow security tools to scan AI ...

Adaptation is essential for survival. Across species, it occurs over many generations through evolution and natural selection ...

Anthropic has accused three Chinese AI firms of running coordinated, large-scale operations to steal capabilities from its Claude models. The U.S.-based company ...