1+ hour, 51+ min ago  (375+ words) A new plug-and-play memory architecture from MIT CSAIL and collaborators could reshape how AI agents handle evolving knowledge, with implications for crypto's growing AI infrastructure layer. The framework is called Me Mo, short for Memory as a Model. It was…...

20+ hour, 25+ min ago  (477+ words) A new formal proof shows that the Joint Embedding Predictive Architecture can reliably recover hidden causes of the world, but only when specific mathematical conditions are met. Yann Le Cun has been talking about world models for years. Now his…...

1+ day, 9+ hour ago  (329+ words) The former Ethereum mining company rolls out secure sandbox environments designed to let AI agents learn from live production scenarios while cutting training costs. Core Weave, the AI infrastructure company that once made its living mining Ethereum, just unveiled a…...

2+ day, 1+ hour ago  (286+ words) Mini Max, the Shanghai-based AI lab backed by Tencent, Alibaba, and mi Ho Yo, just dropped a technical report on its M2 model series. Buried inside was a tease of its next-generation M3 model, which the company claims achieves a 15. 6x faster decoding…...

2+ day, 23+ hour ago  (267+ words) The Di Lo Co X framework achieved 357x better communication efficiency than traditional methods, all over standard 1 Gbps network links. Training a 107-billion-parameter AI model is hard enough when you have a warehouse full of cutting-edge GPUs connected by ultra-fast networking....

3+ day, 2+ hour ago  (509+ words) The Shanghai-based AI lab's latest voice model outperformed GPT Realtime 1. 5 and Gemini Live across all five major benchmarks, introducing paralinguistic comprehension that reads between the lines of human speech. A Shanghai-based AI lab just quietly embarrassed some of the biggest…...

3+ day, 13+ hour ago  (338+ words) The Palo Alto startup says its Axiom Prover system has cracked four long-standing math problems, backed by $200 M in Series A funding and a $1. 6 B valuation. A startup founded by a Stanford dropout less than 15 months ago claims to have…...

3+ day, 23+ hour ago  (429+ words) Google Research wants AI to start saying "I'm not sure" more often. A paper from the company's researchers argues that large language models should hedge their answers when internal confidence is low, rather than delivering every response with the unearned…...

6+ day, 18+ hour ago  (393+ words) The AI system pairs large language models with formal proof-checking to crack decades-old math problems for a few hundred dollars each, and the implications for AI-driven verification stretch far beyond academia. A machine just solved math problems that stumped humans…...

1+ week, 1+ day ago  (657+ words) A general-purpose AI system just overturned a 79-year-old mathematical conjecture by connecting algebraic number theory to plane geometry, and the proof may be headed to the Annals of Mathematics. A general-purpose reasoning model built by Open AI has disproved a…...