About
Cristian Enrique Muñoz Villalobos is a machine learning engineer specializing in deep learning, natural language processing, and high-performance computing. At Holistic AI, he has developed tools for AI risk management, established benchmarks for evaluating AI models, and devised strategies to mitigate AI technical risks. Cristian holds a Master's and Ph.D. in Electrical Engineering from the Pontifical Catholic University of Rio de Janeiro. He has also served as an educator at the Pontifical Catholic University of Peru (PUCP) and the Pontifical Catholic University of Rio de Janeiro (PUC-Rio).
Cristian Enrique Muñoz Villalobos 's articles (3)
This blog explores the evolution of large language models, starting from the launch of InstructGPT, addressing their capabilities, challenges, and limitations, as well as analyzing emerging perspectives and approaches in the field. The content focuses on key AI concepts such as transformers and prompt engineering. It also discusses generative AI models, their potential for artificial general intelligence (AGI), and new challenges and concerns in society regarding their use and development. Finally, innovative strategies are mentioned that seek to overcome current limitations of language models and improve their effectiveness and versatility.
At the heart of large language models lie the innovative Transformer architecture. A deep learning model that has redefined the way we process natural language text due to its remarkable efficiency. In this article, we dive into the details of Transformer, exploring its impressive history of modification and improvement. By the end, you'll have a solid grasp of the cutting-edge technology driving the language models of today.
Speech tech is widely used & has many applications, incl. ASR for voice control of devices & accessing info. However, ASR systems can be fragile & biased, affecting certain groups. This post explores ASR bias metrics used to measure it and recommends datasets to consider.