← Back to feed Tech & Digital

Anthropic found a hidden space where Claude puzzles over concepts

MIT Technology Review 09 July 2026 2h ago
Anthropic found a hidden space where Claude puzzles over concepts
70
Relevance
14/25
Freshness
25/25
Authority
18/20
Brand Signal
10/15
Depth
3/15
Relevance Freshness Authority Brand Depth
The AI firm Anthropic has developed a technique that has given it the clearest glimpse yet at what’s really going on inside large language models as they answer questions or carry out tasks. What they found ranges from the mundane to the unnerving. Researchers at the company built a tool called the Jacobian lens (or…
Read Full Article → MIT Technology Review ↗