So, we will have Claude-Sonnet-5 instead of Fable 5 soon. Looks like a busy week: probably GPT-5.6 and Sonnet 5. But hey, keep it up!
RESEARCH
-
Europe cannot rent its way to AI sovereignty
By
–
Europe cannot rent its way to AI sovereignty. TLDR, here is my opinion that I shared this week with leaders of cutting-edge AI labs. When Washington can disable a model overnight, the question is not whether AI is
-
Fixed Point Reasoners for Stable and Adaptive Looped Transformers
By
–
« Fixed Point Reasoners: Stable and Adaptive Deep Looped Transformers » While Looped Transformers can devote more depth to harder problems, they still need a good method to know when to stop. This article makes of
-
Gary Marcus argues math doesn’t disprove LLM novelty
By
–
this claim around creativity is too strong, IMHO. truly novel ideas from LLMs are surely rare but i don’t think that any math proves they are impossible. note that the objective function and the outputs are not the same.
-
Surprise: GLM-5.2 third behind GPT-5.5 and Opus 4.8
By
–
When I read all the posts about the general surprise at the actual performance of GLM-5.2, which matches the claims, and that many benchmarks confirm it (generally just behind GPT-5.5 and Opus 4.8 in third place), I can even imagine that the founder
-
Turn any paper into running code with autoarxiv
By
–
Turn any paper into running code.
— Akshay 🚀 (@akshay_pachaar) 21 juin 2026
Just swap arxiv → autoarxiv in the paper url.
That hands the paper to an AI agent from alphaXiv. It reads the abstract, the claims, and the linked GitHub repo, then clones the codebase and works through the usual setup pain like dependencies,… pic.twitter.com/UOPJnWdfLJTurn any paper into running code. Just swap arxiv → autoarxiv in the paper url. That hands the paper to an AI agent from alphaXiv. It reads the abstract, the claims, and the linked GitHub repo, then clones the codebase and works through the usual setup pain like dependencies,
-
From RAG retrieval to knowledge compilation: the LLM Wiki
By
–
RAG is already becoming the “old way” The future of AI memory is not retrieval.
It’s compilation. Here’s the shift in one sentence: From searching information To structuring knowledge The new model? LLM Wiki Instead of: Chunking documents Running similarity -
Human Beings as Data: Frontier Labs Lead AI Learning
By
–
Remember, human beings are the data. The more human beings you have on your system, the faster your AI learns. So, who has the most human beings? The Frontier Labs do.
-
Vercel CEO impressed by GLM-5.2, open source and open weights
By
–
Even the CEO of Vercel is impressed/shocked by the exceptional performance of GLM-5.2 in coding. open source, open weights.
-
Do as I Do: Dexerous Manipulation Data from Everyday Human Videos
By
–
“Do as I Do: Dexterous Manipulation Data from Everyday Human Videos” With how robot dexterity is bottlenecked by data as teleoperation and MoCap are expensive and internet videos are only observational, this paper turns normal RGB human videos into executable robot hand