On Bedrock, prompt caching lets open-weight models reuse the long context that generative AI apps otherwise resend with every request, cutting input costs and latency.
This work proposes graph consistency regularization, injecting relational graph structures derived from model predictions into prompts to make prompt optimization more stable and reliable.
At Holiday Extras, OpenAI's dots push out launch reports on their own, point out issues that might slip by, and even talk through upcoming meetings during the commute.
datasette-agent 0.6a0 is Datasette's agent plugin for asking questions about your data, and this release adds an agent prompt directly in the interface.
Your bill can keep climbing even after prompts and app logic are tuned, because four per-request settings — max tokens, reasoning effort, service tier, and cache duration — all move the cost.
Claude Fable 5.2 is sensitive to who owns the code it touches — in Anthropic's small-task tests, it defaulted to building on existing structure when the prompt left ownership unmentioned.