Starting off as a muggle that naïve to the Math's and Data Science world.
Recent Posts
- How to Train Your Deep Research Agent? Prompt, Reward, and Policy Optimization in Search-R1ref: https://arxiv.org/pdf/2602.19526v1 Summary I used to think longer prompts and step-by-step reasoning always make AI smarter. In daily use, that often helps. But in reinforcement learning, the opposite… Read more: How to Train Your Deep Research Agent? Prompt, Reward, and Policy Optimization in Search-R1
- Prompt Repetition Improves Non-Reasoning LLMs
ref: https://arxiv.org/pdf/2512.14982 Summary The easiest AI upgrade might be just “copy and paste”. Google researchers tested with 7 leading models and found that repeating the exact same prompt… Read more: Prompt Repetition Improves Non-Reasoning LLMs - Verbalized Sampling: How to mitigate Mode Collapse and unlock LLM diversityref: https://arxiv.org/pdf/2510.01171 Summary AI models repeat safe answers, but a simple prompt tweak can unlock far better thinking. Researchers introduced “verbalized sampling,” where the model lists possible answers… Read more: Verbalized Sampling: How to mitigate Mode Collapse and unlock LLM diversity
- Tool-MAD: A Multi-Agent Debate Framework for Fact Verification with Diverse Tool Augmentation and Adaptive Retrieval
ref: https://arxiv.org/pdf/2601.04742 Summary The team studied several methods to improve LLM accuracy and combined the strongest ideas to build Tool-MAD. The system uses evolving AI debaters that run… Read more: Tool-MAD: A Multi-Agent Debate Framework for Fact Verification with Diverse Tool Augmentation and Adaptive Retrieval - The Next Conversation
Understanding is a superpower. Liking is optional. Jefferson Fisher teaches a simple but powerful idea: Argue less. Talk more The book changed the way i communicate. 7 Powerful… Read more: The Next Conversation
Site Links
Kim 2 ML 