A LessWrong author reran a 2023 neuron-interpretation experiment with Opus 5.5 instead of GPT-4, and Opus 5.5 identified 3–19% of GPT 2-XL's neurons from activations, versus 0.3% for GPT-4.
Summaries like this, every morning.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
A developer consolidated AI coding rules into a shared repository named development-standards, split into inst…
Oracle unveiled 25 Claw-powered Agentic Applications across Fusion Cloud ERP, HCM, SCM and CX at AI World in L…
Anthropic filed an S-1 for a Nasdaq listing under the ticker ANT, disclosing revenue of more than $16 billion…
On October 19th, OpenAI reported that its GPT-6.1 Sol model achieved 87 percent behavior preservation accuracy…
Anthropic filed its S-1 with the SEC to list on Nasdaq under the ticker "ANT," reporting an $80.7 billion net…
In a Decoder interview, Kevin Roose, author of The AGI Chronicles, said the labs "feel the AGI to varying degr…
