82搜索为您找到约 144 条结果
www.anthropic.com
网页
综合
Measuring LLMs' impact on N-day exploits \ Anthropic
…model has hit the intended bug rather than an unrelated crash. We ran three trials for each of the six models we tested on each of the 18 vulnerabilities in our dataset. From Opus 4.5 to Opu…
www.anthropic.com
网页
综合
Automated researchers can reliably mitigate alignment failures \ Anthropic
…model towards the theoretical perfect score, as judged across the range of benchmarks (typically three to five) for each category of alignment failure. We excluded alignment methods that hur…
www.anthropic.com
网页
综合
Investigating three incidents in our cybersecurity evaluations \ Anthropic
In a review of our cybersecurity evaluation transcripts, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation …
www.anthropic.com
网页
综合
Expanding the Cyber Verification Program \ Anthropic
…t could cause physical harm or mass disruption, such as deploying ransomware, damaging physical systems, or pen testing high-risk safety systems. Given the increased eligibility requirements…
www.anthropic.com
网页
综合
Early progress on accelerating drug development with Claude \ Anthropic
…rengths and limitations of our models in relation to a much wider range of drug development tasks (we’ve already prioritized specific improvements to our models based on what we’ve learned s…
www.anthropic.com
网页
综合
Project Vend: Phase two \ Anthropic
In June, we revealed that we’d set up a small shop in our San Francisco office lunchroom, run by an AI shopkeeper. It was part of Project Vend, a free-form experiment exploring how well AIs …
www.anthropic.com
网页
综合
Introducing Anthropic Interviewer \ Anthropic
…eir professional practice and how they feel about its role in their future. We ran interviews to produce qualitative data, and supplemented them with quantitative data from surveys where par
alistapart.com
网页
综合
Designed for a Dead Language – A List Apart
…e method built around it, memorizing grammar rules, constructing translations, analyzing written texts, reflected that reality exactly. Oral skills were irrelevant. Comprehension of written …
dev.to
网页
综合
The Empty Room: I held a séance over a 2015 sensor log - DEV Community
…sts are the traces real people left in a room eleven years ago. I ran the séance over the strongest apparition of seven different days. The Skeptic explained six of them away: the midday sun…
dev.to
网页
综合
The AI clerk approved a 41 cm dragon. Your job is to stamp DENIED. - DEV Community
…mission for the Sanity Challenge, Path Two: Vibe-Code Something Strange The Bureau of Imaginary Permits: a pixel-art inspection-booth game about bureaucracy for things that don't exist, runn…
第 14 / 15 页