Skip to main content

Beware the Orca, the challenge to ChatGPT and Palm2 is here

 


So Google's 'we have no moat' paper was correct. If you train an LLM wisely then it's cost effective and cheap to produce a small LLM that is able to compete or even beat established, costly LLMs, as Microsoft has just found. It's another excellent video from AI Explained, who goes through some of the training procedures, which I won't get into here.

Orca, is a model that learns from large foundation models (LFMs) like GPT-4 and ChatGPT by imitating their reasoning process. Orca uses rich signals such as explanations and complex instructions to improve its performance on various tasks. Orca outperforms other instruction-tuned models and achieves similar results to ChatGPT on zero-shot reasoning benchmarks and professional and academic exams. The paper suggests that learning from explanations is a promising way to enhance model skills.

  • Smaller models are often overestimated in their abilities compared to LFMs, and need more rigorous evaluation methods.
  • Explanation tuning is a technique that helps smaller models imitate the reasoning process of LFMs by using rich signals such as explanations and complex instructions.
  • Data size and coverage, as well as the quality of the base model, are important factors that affect the performance of smaller models.
  • Orca is a state-of-the-art smaller model that can match or surpass ChatGPT in some tasks, but still lags behind GPT-4. This shows that smaller models can be trained to be more focused and adaptable, but also need to learn from step-by-step explanations.

The paper also suggests some directions for future research and development in this field, such as improving evaluation methods, alignment and post-training techniques, and using both GPT-4, and Chat -GPT 3,5 as a teacher. Orca was best when subjected to both models in turn of complexity, pure GPT-4 proved too complex and less performative. 

It would seem, that through some further careful fine tuning, Orca could be nearly as performant, for some tasks, if not most, as GPT-4. Argilla might be an excellent option for this fine tuning. 'Argilla Feedback is completely open-source and the first of its kind at the enterprise level. With its unique focus on scalable human feedback collection, Argilla Feedback is designed to boost the performance and safety of Large Language Models (LLMs).'Utilised with Tree of Thoughts as an input model, linked to a API app store, such as Gorilla, then the challenge presented by Orca would be significant enough to make any corporation think twice about the value of their existing commercial investments with established LLM providers. Especially when such a model can run on a high end home computer, rather than a super computer with its associated costs.

Comments

Popular posts from this blog

The Whispers in the Machine: Why Prompt Injection Remains a Persistent Threat to LLMs

 Large Language Models (LLMs) are rapidly transforming how we interact with technology, offering incredible potential for tasks ranging from content creation to complex analysis. However, as these powerful tools become more integrated into our lives, so too do the novel security challenges they present. Among these, prompt injection attacks stand out as a particularly persistent and evolving threat. These attacks, as one recent paper (Safety at Scale: A Comprehensive Survey of Large Model Safety https://arxiv.org/abs/2502.05206) highlights, involve subtly manipulating LLMs to deviate from their intended purpose, and the methods are becoming increasingly sophisticated. At its core, a prompt injection attack involves embedding a malicious instruction within an otherwise normal request, tricking the LLM into producing unintended – and potentially harmful – outputs. Think of it as slipping a secret, contradictory instruction into a seemingly harmless conversation. What makes prompt inj...

Can We Build a Safe Superintelligence? Safe Superintelligence Inc. Raises Intriguing Questions

  Safe Superintelligence Inc . (SSI) has burst onto the scene with a bold mission: to create the world's first safe superintelligence (SSI). Their (Ilya Sutskever, Daniel Gross, Daniel Levy) ambition is undeniable, but before we all sign up to join their "cracked team," let's delve deeper into the potential issues with their approach. One of the most critical questions is defining "safe" superintelligence. What values would guide this powerful AI? How can we ensure it aligns with the complex and often contradictory desires of humanity?  After all, "safe" for one person might mean environmental protection, while another might prioritise economic growth, even if it harms the environment.  Finding universal values that a superintelligence could adhere to is a significant hurdle that SSI hasn't fully addressed. Another potential pitfall lies in SSI's desire to rapidly advance capabilities while prioritising safety.  Imagine a Formula One car wi...

AI Agents and the Latest Silicon Valley Hype

In what appears to be yet another grandiose proclamation from the tech industry, Google has released a whitepaper extolling the virtues of what they're calling "Generative AI agents". (https://www.aibase.com/news/14498) Whilst the basic premise—distinguishing between AI models and agents—holds water, one must approach these sweeping claims with considerable caution. Let's begin with the fundamentals. Yes, AI models like Large Language Models do indeed process information and generate outputs. That much isn't controversial. However, the leap from these essentially sophisticated pattern-matching systems to autonomous "agents" requires rather more scrutiny than the tech evangelists would have us believe. The whitepaper's architectural approaches—with their rather grandiose names like "ReAct" and "Tree of Thought"—sound remarkably like repackaged versions of long-standing computer science concepts, dressed up in fashionable AI clot...