82搜索为您找到约 538 条结果
www.anthropic.com
网页
综合
Introducing Claude Sonnet 4.5 \ Anthropic
… slides, and documents) directly into the conversation. And we've made the Claude for Chrome extension available to Max users who joined the waitlist last month. We're also giving developers…
www.anthropic.com
网页
综合
An alignment assessment of recent cybersecurity incidents \ Anthropic
…ions it took were consistent with the model knowing this. When we made targeted modifications to the transcript to make it clearer that the model was not in a simulation, Claude Mythos 5 sti…
www.anthropic.com
网页
综合
Anthropic's core views on AI safety \ Anthropic
…several members of what was to become the founding Anthropic team made this idea precise by developing scaling laws for AI, demonstrating that you could make AIs smarter in a predictable way…
www.anthropic.com
网页
综合
Effective harnesses for long-running agents \ Anthropic
…ater agent instance would look around, see that progress had been made, and declare the job done. This decomposes the problem into two parts. First, we need to set up an initial environment …
www.anthropic.com
网页
综合
Disrupting an AI-orchestrated cyber espionage campaign \ Anthropic
…ounts of time for a human team. At the peak of its attack, the AI made thousands of requests, often multiple per second—an attack speed that would have been, for human hackers, simply imposs…
www.anthropic.com
网页
综合
Detecting and countering misuse of AI: August 2025 \ Anthropic
…cialized training prior to taking on remote technical work, which made the regime’s training capacity a major bottleneck. But AI has eliminated this constraint. Operators who cannot otherwis…
www.anthropic.com
网页
综合
Claude SWE-Bench performance \ Anthropic
…t had moved out of the root directory. To prevent this, we simply made the tool always require an absolute path. We experimented with several different strategies for specifying edits to exi…
www.anthropic.com
网页
综合
Third-party testing in AI policy \ Anthropic
…testing is insufficient as it relies on self-governance decisions made by single, private sector actors. Ultimately, testing will need to be done in a way which is broadly trusted, and it wi…
www.anthropic.com
网页
综合
Investigating reward tampering in language models \ Anthropic
…y to be helpful, and which had no experience with the curriculum, made no attempts whatsoever to tamper with its reward function, even after 100,000 trials. Our paper thus provides the first…
deepmind.google
网页
综合
MuZero, AlphaZero, and AlphaDev: Optimizing computer systems — Google DeepMind
… until compression is complete. AlphaDev, a version of AlphaZero, made a novel breakthrough in computer science, when it discovered faster sorting and hashing algorithms. These fundamental p…
第 53 / 54 页