Gustavo de Rosa
Member of Technical Staff
关于
As a Member of Technical Staff in the AI Frontiers lab within Microsoft Research, Redmond, USA, I am privileged to be part of such an esteemed organization and to contribute to research that has real-world impact.
My journey in the field of computer science began during my undergraduate studies, where I was initially uncertain of the path to pursue. However, through my Bachelor of Computer Science at Universidade Estadual Paulista Júlio de Mesquita Filho (opens in new tab), FC/Bauru (2016) and participation in a research laboratory, I discovered my passion for artificial intelligence trends. This led me to pursue further academic endeavors and attain a Master of Science in Computer Science at Universidade Estadual Paulista Júlio de Mesquita Filho, IBILCE/Rio Preto (2018) where I gained expertise in image processing formulations, pattern recognition, pattern classification, machine learning algorithms, and meta-heuristic optimization.
I achieved my Ph.D. at Universidade Estadual Paulista Júlio de Mesquita Filho, FC/Bauru (2022) with a focus on natural language processing and adversarial learning, further solidifying my expertise in the field. I am excited to continue my professional growth and contribute to the advancement of technology at Microsoft Research.
精选内容
Phi-4-reasoning Technical Report
We introduce Phi-4-reasoning, a 14-billion parameter reasoning model that achieves strong performance on complex reasoning tasks. Trained via supervised fine-tuning of Phi-4 on carefully curated set of “teachable” prompts–selected for the right level of complexity and diversity–and reasoning demonstrations generated…
Phi-4 Technical Report
We present phi-4, a 14-billion parameter language model developed with a training recipe that is centrally focused on data quality. Unlike most language models, where pre-training is based primarily on organic data sources such as web content or code, phi-4…
Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
We introduce phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion tokens, whose overall performance, as measured by both academic benchmarks and internal testing, rivals that of models such as Mixtral 8x7B and GPT-3.5 (e.g., phi-3-mini achieves 69%…
Phi-2: The surprising power of small language models
Phi-2 is now accessible on the Azure model catalog. Its compact size and new innovations in model scaling and training data curation make it ideal for exploration around mechanistic interpretability, safety improvements, and fine-tuning experimentation on a variety of tasks.
Textbooks Are All You Need
We introduce phi-1, a new large language model for code, with significantly smaller size than competing models: phi-1 is a Transformer-based model with 1.3B parameters, trained for 4 days on 8 A100s, using a selection of ``textbook quality" data from…
Archai: Platform for Neural Architecture Search
Archai accelerates your Neural Architecture Search (NAS) through fast, reproducible and modular research, allowing you to generate efficient deep networks for your applications.