Insane New AI Model - ORCA - That Finally Beats GPT-4

Microsoft has released a new research paper on Orca, a revolutionary AI model that learns from complex explanations of GPT4, the most powerful language model in the world. This is a huge deal and I'm going to tell you why in this video. So what is Orca and why is it so important? Well, ORCA is a 13 billion parameter model that learns from complex explanation traces of GPT4, which is a much bigger model than that, can generate almost any kind of text you can imagine. Now why would Microsoft want to create a smaller model that learns from a bigger model? Isn't bigger always better when it comes to AI? Well, not exactly. Bigger models are more powerful, but they also have some drawbacks. They are very expensive to train and run, they require a lot of computing resources and energy, and they are not very accessible to most researchers and developers. That's why there has been a lot of interest in creating smaller models that can still perform well on various tasks, such as answering questions, summarizing texts, generating captions, and so on. These smaller models are usually fine tuned on specific data sets or instructions to make them more specialized and efficient. However, there is a problem with this approach. Smaller models tend to have poor reasoning and comprehension skills compared to bigger models. They often make mistakes or give irrelevant answers when faced with complex or ambiguous queries. They also lack the ability to explain how they arrived at their answers or what steps they took to solve a problem. But ORCA is not just another smaller model that imitates a bigger model. ORCA is a smaller model that learns from the reasoning process of a bigger model. It learns from the explanations that GPT4 gives when it generates its answers. These explanations are not just simple sentences or phrases. They are detailed traces of how GPT4 thinks step by step. How it uses logic and common sense, how it connects different pieces of information, and how and how it simplifies complex concepts. By learning from these explanations, ORCA becomes much more capable and intelligent than other models. It can handle more diverse and challenging tasks. It can give more accurate and relevant answers, and it can also explain its own reasoning process to humans. This is a huge breakthrough for open source AI. ORCA is set to be open source soon, which means anyone will be able to use it and build upon it. It will enable more people to access the power of GPT4 without having to pay for it or deal with its limit. ORCA will also open up new possibilities for AI research and development, especially in areas that require more reasoning and understanding skills. To understand how ORCA works, we need to first understand how GPT4 works. So GPT4 is more than a text generator. It performs tasks requiring reasoning like answering factual questions, summarizing lengthy texts, generating captions, writing essays, and more. Interestingly, GPT4 can provide explanations for its outputs. These are found in the model's internal states, essentially its thoughts or memories, which hold the logic and information used to generate outputs. By using specific prompts, we can unveil these internal explanations, giving a detailed view of how GPT4 thinks, solves problems, and uses diverse sources of information, including its own memory, the web, and common sense. These explanations are very valuable for smaller models that want to learn from GPT4. They provide more signals and guidance for how to perform various tasks and how to reason and understand different concepts. They also make the learning process more transparent and interpretable for humans. This is what ORCA does. ORCA learns from these explanations that GPT4 generates when it performs different tasks. It uses these explanations as its training data and tries to imitate them as closely as possible. ORCA also tries to generate its own explanations when it performs similar tasks and compares them with GPT4's explanations to improve itself. So ORCA is actually based on Vicuna, a previous open source model that was fine tuned on question answer pairs from GPT 3.5. Orca extends Vicuna by adding a new technique called explanation tuning, which allows it to learn from complex explanation traces of GPT4. Explanation tuning is a fresh approach that enhances GPT4's skill to follow specific directives. By refining this AI with prompts like Summarize this in a sentence or Create a love haiku, we make it more adept at particular tasks. But explanation tuning goes beyond it hones GPT4 to reveal its thought process, using prompts like Think sequentially or Explain like I'm a child. This way, GPT4's reasoning becomes more transparent. This technique involves standard and explanation prompts. The former are usual tasks like who leads France? Or Craft a winter poem. The latter instruct GPT4 to clarify its logic, like Think in steps or show how you did it. Using both prompt types together, GPT4 produces complex explanation traces. For instance, with the standard prompt who leads France? And the explanation prompt think in steps, GPT4 might provide a step by step explanation. This comprehensive response not only tells us who the president is, but also illustrates GPT4's problem solving strategy and information sources, offering more insight than a simple answer. ORCA leverages explanation traces as learning material, striving to mimic them and generate its own for improvement. But where do these traces come from Orca taps into Flan 2022, a massive collection of over 1,000 tasks and 10,000 instruct covering a spectrum of subjects. By sampling from Flan 2022, Orca gets a variety of tasks and uses them to query GPT4 for explanation traces. It also creates complex prompts from the data set to test GPT4's reasoning, like mashing two tasks into one. This way, Orca learns from diverse and intricate tasks, fostering many aspects of human intelligence. ORCA is evaluated on a number of benchmarks that test its generative reasoning and comprehension abilities. These benchmarks include multiple choice questions, natural language inference, text summarization, text generation, image captioning, and so on. ORCA is compared with other models of similar size or larger size, such as Vicuna 13B, text DAVINCI003, a free version of GPT3, chat GPT 3.5 and GPT4. Orca's performance is stellar, topping all other open source models in most benchmarks, particularly those needing deeper reasoning. Despite its smaller size, it matches or beats ChatGPT in many areas, even competing with GPT4 in tasks like natural language inference or image captioning. Here's a quick look at Orca's benchmark performances. On BigBench hard, it scores a 64% accuracy, more than double of Vicuna 13B's and surpassing ChatGPT's 59% and GPT4s. On Superglue, it achieves an 86% average, beating Vicuna 13B. Text DaVinci003 83%, ChatGPT 84% and nearly matching GPT4. On CNN Daily Mail, Orca earns a rougel score of 41%, outperforming Vicuna 13B. Text DaVinci003 39%, ChatGPT 40% and closing in on GPT4 42%. On Coco captions, it scores a cider of 120% higher than Vicuna 13B, 113%, text DaVinci003 115%, ChatGPT 117% and GPT4 119%. So as you can see, Orca is a highly versatile, efficient model, performing well across tasks and domains and soon to be open source. It also works on a single GPU. Orca's success reveals multiple insights about AI's future. Firstly, it indicates that learning from explanations as opposed to just answers notably boosts AI intelligence and performance. By studying GPT4's explanations, Orca not only gains superior reasoning skills but also provides a transparent look into its problem solving process. Secondly, ORCA proves that despite their size, smaller models can match or outperform larger ones. Learning from GPT4, Orca sidesteps size related drawbacks, showing that smaller models can be more approachable and efficient, needing fewer resources and energy. And thirdly, ORCA exemplifies how open source AI through inventive methods can match proprietary AI and demonstrates how open source AI's wider accessibility can benefit more people and spur more applications. Concerning its positioning, ORCA isn't just a mini GPT4 or another open source model. While it doesn't match GPT4's broad capacity or knowledge base, it harnesses GPT4's reasoning, making it smarter than other small models. It also surpasses GPT4 in transparency by generating its own explanation traces. Unlike other open source models, ORCA learns from a varied range of tasks and complex explanations, making it more intelligent and versatile. Therefore, ORCA occupies a unique position in the AI sphere, combining GPT4's prowess with open source AIs accessibility and demonstrating the potential of explanation based learning alright, that's it for this video. Thank you so much for watching. If you liked this video, please give it a thumbs up and share it with your friends. And if you haven't already, please subscribe to my channel and turn on the notification bell so you don't miss any of my future videos. I'll see you in the next one.