ai

to see

• what is gen ai ibm video

  llms

• ref
• Large Language Models have been trained with massive amounts of data. These models have billions of parameters which are the weights of the connections between neurons in the neural networks
• The more closely neurons are related, the bigger the weights between them are
• Collections of neurons that related to each other are clustered together. For example, there can be clusters like ‘recipes’, ‘flowers’, ‘English’ and many more.
• Prompt Engineering helps the AI to find the right entry points in these huge networks. For example, if you look for ingredients for a recipe, good prompts help the AI to start looking for answers in the ‘recipes’ cluster rather than in the ‘flowers’ cluster.
• n addition to the text of prompts often parameters can be tuned. The ‘temperature’ defines whether the model should try to be as precise as possible or whether it should be as creative as possible. Via other parameters the lengths of the outputs can be defined.

  gen ai

• ref
• Generative AI or generative artificial intelligence refers to the use of AI to create new content, like text, images, music, audio, and videos.
• enerative AI is powered by foundation models (large AI models) that can multi-task and perform out-of-the-box tasks, including summarization, Q&A, classification, and more. Plus, with minimal training required, foundation models can be adapted for targeted use cases with very little example data.
• Generative AI works by using an ML model to learn the patterns and relationships in a dataset of human-created content. It then uses the learned patterns to generate new content.

  prompt engineering

• pdf ref
• ref
• Prompt engineering is a relatively new discipline for developing and optimizing prompts to efficiently use language models (LMs) for a wide variety of applications and research topics.
• Prompt engineering skills help to better understand the capabilities and limitations of large language models (LLMs).

  few-shot prompts
  • Chain-of-thought (CoT) prompting
  • Self-Consistency
  • Knowledge Generation Prompting
  • ReAct
  

  Model safety
  • Prompt Injection
  • RLHF:Reinforcement Learning from Human Feedback (RLHF) is now being used to train LLMs that fit human preference  
  Popular examples:
  • Claude (Anthropic)
  • ChatGPT (OpenAI)
  • Future dir
  ections
  Prompt Injection
  • Prompt Leaking
  • Jailbreakin