Architecture and building blocks of Agentic Systems | What Antropic says on the topic

Antropic article https://www.anthropic.com/research/building-effective-agents https://github.com/anthropics/anthropic-cookbook/tree/main/patterns/agents VERDICT: A Library for Scaling Judge-Time Compute https://arxiv.org/pdf/2502.18018 Agent Example 1: Raising the bar on SWE-bench Verified with Claude 3.5 Sonnet https://huggingface.co/datasets/princeton-nlp/SWE-bench_Lite https://github.com/SWE-bench/SWE-bench https://swe-agent.com/latest/ https://github.com/SWE-agent/SWE-agent/blob/main/docs/usage/hello_world_output.txt Agent example 2: Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku https://www.anthropic.com/research/swe-bench-sonnet https://www.anthropic.com/news/3-5-models-and-computer-use https://docs.anthropic.com/en/docs/agents-and-tools/computer-use https://www.youtube.com/watch?v=ODaHJzOyVCQ&ab_channel=Anthropic Agent example 3: OpenaAI Operator https://help.openai.com/en/articles/10421097-operator Understanding the planning of LLM agents: A survey https://arxiv.org/pdf/2402.02716