Shahar Keren

Title: Shahar Keren: Innovator in Language Models and Recommendation Systems

Introduction

Shahar Keren is a notable inventor based in Hemed, Israel. He has made significant contributions to the fields of language models and recommendation systems, holding a total of 5 patents. His work focuses on enhancing the accuracy and efficiency of language models, which are increasingly vital in various applications.

Latest Patents

One of Shahar Keren's latest patents is titled "Dynamic construction of large language model prompts." This invention provides example solutions for reducing the likelihood of hallucinations by language models, such as large language models (LLMs). By injecting a sufficient range and quantity of curated factual data into a prompt, the likelihood of hallucination by an LLM may be reduced. This innovation enables language models to be utilized in a wider range of settings, where the fabrication of facts is problematic, while minimizing the need for human verification of generated text.

Another significant patent is the "Multilingual content-based recommendation system." This invention outlines methods and apparatuses that access multiple sources of information concerning features for applications. It cleans the data from these sources, extracts features, selectively weights the sources or extracted features, and produces a feature vector. This feature vector can be used in either a single language feature space or a multi-language feature space, facilitating the recommendation of applications based on similarities found through a connected graph of feature spaces.

Career Highlights

Shahar Keren is currently employed at Microsoft Technology Licensing, LLC, where he continues to innovate and develop new technologies. His work at Microsoft has allowed him to collaborate with other talented individuals in the field, contributing to advancements in technology and research.

Collaborations

Some of Shahar's coworkers include Nir Nice and Noam Koenigstein. Their collaborative efforts have likely played a role in the successful development of innovative solutions in their respective fields.

Conclusion

Shahar Keren is a prominent inventor whose work in language models and recommendation systems is paving the way for more accurate and efficient technologies. His contributions are significant in enhancing the capabilities of language models and improving user experiences in application recommendations.