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

Just as general-purpose models opened the era of practical AI, narrow, orchestrated models could define the economics and ...

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

In many real-world systems, such as fireflies flashing, Japanese tree frogs croaking, magnetic Janus particles, neurons firing, and in various engineered systems like drone swarms and wireless sensor ...

People's decisions are known to be influenced by past experiences, including the outcomes of earlier choices. For over a century, psychologists have been trying to shed light on the processes ...

OpenAI released GPT-5.4 today with native computer use, a 1M-token context window, and new professional benchmarks. Find what ...

Databricks' KARL agent uses reinforcement learning to generalize across six enterprise search behaviors — the problem that breaks most RAG pipelines.

A preprint paper submitted to arXiv on Jan. 22, 2026, ranks common chickens higher than leading AI systems on a new consciousness scoring framework, placing the humble barnyard bird above models like ...

OpenAI has launched GPT-5.4, its latest frontier AI model designed for professional workloads, offering a massive one-million-token context window, improved reasoning, lower error rates and faster ...

Ten AI concepts to know in 2026, including LLM tokens, context windows, agents, RAG, and MCP, for building reliable AI apps.

"Glass" has a unique and distinct meaning in physics—one that refers not just to the transparent material we associate with window glass. Instead, it refers to any system that looks solid but is not ...

Late in 2025, we covered the development of an AI system called Evo that was trained on massive numbers of bacterial genomes. So many that, when prompted with sequences from a cluster of related genes ...