o1-preview: OpenAI's New AI Model that can Think & Reason ??

Welcome to the?latest AI in 5?newsletter with Clarifai!

Every week we bring you new models, tools, and tips to build production-ready AI!

Here is the summary of what we will be covering this week: ??

• New models: o1-preview and o1-mini from OpenAI
• Webinar: AI for marketers - Content generation and personalization with visual search
• Tutorial: Multistage RAG pipeline with DSPy
• Tip of the week: Positive and negative annotations

New Models from OpenAI ??

OpenAI has released o1, a new series of AI models that use reinforcement learning to enhance reasoning abilities.

These models, including o1-preview and o1-mini, employ reasoning tokens to process complex problems before generating responses.

Here are some key capabilities:

• Unlike its predecessors, which prioritized pre-training, o1 invests in extended inference, sharpening its logical abilities.
• The model scored in the 89th percentile on Codeforces programming competitions and 83% on International Mathematics Olympiad qualifying questions, compared to GPT-4o's 14%.
• It can solve PhD-level problems in physics, biology, and chemistry.
• o1-preview, designed for refining use cases, and o1-mini, a smaller model that is 80% cheaper than o1-preview and delivers almost the same performance for STEM challenges.

The models are now available on the Clarifai Platform. Try out o1-preview and o1-mini, and access them via the API! ??

AI for Marketers: Content Generation and Personalization with Visual Search ??

We are hosting a webinar next week!?

Most organizations struggle with building and deploying an AI framework that helps marketers find and deliver the right content at the right stage of the buyer's journey.

Join us for this webinar where we will explore the power of AI-based content organization and personalization, tailored specifically for leaders in marketing, retail, and e-commerce.?

Here’s what you will learn:

• Why and how to use AI for personalized content and product recommendations
• How to organize content by tagging and enriching data, and leverage AI models to create personalized content strategies
• How to create a common language and framework for AI development to enhance marketing strategies

Register below??

Content Generation and Personalization with Visual Search

Multistage RAG Pipeline with DSPy ??

Retrieval-Augmented Generation (RAG) is a technique that enhances language models by combining them with external knowledge retrieval.?

DSPy is a framework for solving advanced tasks using language models and retrieval models. It offers the flexibility to algorithmically optimize language model prompts and weights.?

The Multistage RAG pipeline uses DSPy to create a multi-stage retrieval system. It:

1. Extract keywords based the user query to expand the search scope
2. Retrieves information using both the user query and keywords
3. Uses a separate language model to rank passages by relevance

This multi-stage approach enhances the retrieval accuracy and comprehensiveness.

Check out the below tutorial: Multistage RAG with DSPy

Tip of the Week: ??

Positive and Negative Annotations!

Positive and negative annotations are essential for training machine learning models. Positive annotations help the model learn relevant features, while negative annotations refine differentiation by showing what to exclude.?

Using both types of annotations reduces misclassification and enhances model accuracy and reliability.

Below is an example of adding both positive and negative annotations to your input ID in the Clarifai app using a cURL request. Check out the complete guide here.

Want to learn more from Clarifai? “Subscribe” to make sure you don’t miss the latest news, tutorials, educational materials, and tips. Thanks for reading!