Project Fetch: Can Claude train a robot dog?
…Can Claude train a robot dog? Nov 12, 2025 How could frontier AI models like Claude reach beyond computers and affect the physical world? One path is through robots. We ran an…
Search
People also ask
What is a natural language autoencoder?
The core idea is to train Claude to explain its own activations. But how do we know whether an explanation is good? Since we don't know what thoughts an activation actually encodes, we can't directly check whether an explanation is accurate. So we train a second copy of Claude to work backwards—reconstruct the original activation from the text explanation. We consider an explanation to be good if it leads to an accurate reconstruction. We then train Claude to produce better explanations according to this definition using standard AI training techniques. In more detail, suppose we have a langua
Natural Language AutoencodersClaude for Financial Services
…Tailored onboarding, training, and best practices for rapid value realization. Financial institutions require the highest standards of data protection. By default, your data is not used for training our generative models, maintaining…
Emergent introspective awareness in large language models
…Why does this matter? We think understanding introspection in AI models is important for several reasons. Practically, if introspection becomes more reliable, it could offer a path to dramatically increasing the transparency…
2028: Two scenarios for global AI leadership
…With the supply of compute expanding rapidly, and with AI being used increasingly to augment the training of new AI models, we’re entering a period of great acceleration in AI capabilities…
TCS and Anthropic partner to bring Claude to regulated industries
…TCS iON , which conducts more than 75 million assessments each year across 1,500 cities in India, will deliver Claude training and certification. “Enterprise AI value comes from understanding business context, orchestrating…
Emotion concepts and their function in a large language model
…What’s behind these behaviors? The way modern AI models are trained pushes them to act like a character with human-like characteristics. In addition, these models are known to develop rich…
Automated Alignment Researchers: Using large language models to scale scalable oversight
…Aligning smarter-than-human AI models is a research area known as “scalable oversight”. Scalable oversight has largely been discussed in theoretical, rather than practical , terms—but at AI’s current pace…
KPMG integrates Claude across its core business and workforce of more than 276,000 in strategic alliance
…modernize aging IT systems faster and ship new AI-enabled technology in a fraction of the usual time. “Human in the loop” in practice Joint research from KPMG and the McCombs School…
Core views on AI safety: When, why, what, and how
…In our AI safety research, empirical evidence about AI – though it mostly arises from computational experiments, i.e. AI training and evaluation – is the primary source of ground truth. This doesn’t…
Building AI for cyber defenders
Policy Frontier Red Team Building AI for cyber defenders Oct 3, 2025 AI models are now useful for cybersecurity tasks in practice, not just theory. As research and experience demonstrated the utility…