- Talk "Deep Reinforcement Learning in the Real World: From Chip Design to LLMs"Link visible for attendees
To access this session, please register here: https://hubs.li/Q02rF3Gl0
Topic: "Deep Reinforcement Learning in the Real World: From Chip Design to LLMs"
Speaker: Anna Goldie, Senior Staff Research Scientist at Google DeepMind
Anna works on Large Language Model (LLM) research in Gemini & Bard. Previously, she worked on RL for LLMs and retrieval-augmented LLMs at Anthropic and was co-founder/lead of the ML for Systems team in Google Brain. Her RL methods have been used in multiple generations of Google's flagship AI accelerator (TPU). She graduated from MIT with a Bachelors in Computer Science, a Bachelors in Linguistics, and a Master of Computer Science, and is a CS PhD Candidate in the Stanford NLP Group. She has published peer-reviewed articles in top scientific venues, including Nature, NeurIPS, ICLR, EMNLP, ISPD, ASPLOS, and MLCAD. She was named one of MIT Technology Review's 35 Innovators Under 35, and her work has been covered in various media outlets, including CNBC, IBTimes, IEEE Spectrum, MIT Technology Review, WIRED. and ABC News.
Abstract:
Reinforcement learning (RL) is famously powerful but difficult to wield, and until recently, had demonstrated impressive results on games, but little real world impact. I will start the talk with a discussion of RL for Large Language Models (LLMs), including scalable supervision techniques to better align models with human preferences (Constitutional AI / RLAIF). Next, I will discuss RL for chip floorplanning, one of the first examples of RL solving a real world engineering problem. This learning-based method can generate placements that are superhuman or comparable on modern accelerator chips in a matter of hours, whereas the strongest baselines require human experts in the loop and can take several weeks. This method was published in Nature and used in production to generate superhuman chip layouts for the last four generations of Google’s flagship AI accelerator (TPU).Hybrid ODSC East 2024 on 23rd-25th April - https://hubs.li/Q027_nYw0
Use COMMUNITY-EAST2024 - code for extra discount on any pass of your choice.ODSC Links:
• Get free access to more talks/trainings like this at Ai+ Training platform:
https://hubs.li/H0Zycsf0
• ODSC blog: https://opendatascience.com/
• Facebook: https://www.facebook.com/OPENDATASCI
• Twitter: https://twitter.com/_ODSC & @odsc
• LinkedIn: https://www.linkedin.com/company/open-data-science
• Slack Channel: https://hubs.li/Q02r9VZM0
• Code of conduct: https://odsc.com/code-of-conduct/ - Enhancing Security Report Generation with RAG and Fine-Tuned Language ModelsLink visible for attendees
To access this webinar, please register here: https://hubs.li/Q02rZ7s70
Topic: "Enhancing Security Report Generation with RAG and Fine-Tuned Language Models: Integrating Account Telemetry and OCI Cloud Guard"
Speaker: Dr. Sanjay Basu, Senior Director – AI/ML at Oracle Cloud Engineering
Dr. Sanjay Basu is an industry-recognized subject matter expert in Artificial Intelligence, Machine Learning, and Quantum Computing. He has double Master’s in computer science and systems design. His PhD was in Organizational Behaviour and Applied Neuroscience. Currently, he is pursuing his second PhD in AI with focus of research in Retentive Networks. Dr. Basu is also the author and editor of Ethics in AI collection, author of Web 3 books. View his latest blogs here.Abstract:
In the rapidly evolving landscape of cyber security, the ability to swiftly generate comprehensive and accurate security reports is paramount. This session showcases an advanced approach that leverages Retrieval-Augmented Generation (RAG) and fine-tuned Large Language Models (LLMs), such as Cohere and Llama2, to automate the creation of detailed security reports. By incorporating account telemetry and network traffic logs as RAG, this method enhances the report's contextuality and relevance, ensuring precise and insightful incident narratives and breach analyses.
A focal point of the presentation is the innovative integration of logs from OCI Cloud Guard, a key component in identifying and correlating security threats within the Oracle Cloud Infrastructure (OCI). This integration enriches the reports with specific cloud-based security insights and leverages OCI's advanced threat detection capabilities to provide a more granular understanding of the security posture.
The session discusses the integration of Oracle Database 24c's vector search capabilities into the solution, significantly improving the efficiency and accuracy of data retrieval processes involved in report generation. This cutting-edge database feature enables rapid searching of large volumes of data, including unstructured data such as logs and incident reports, by using vector embeddings to find the most relevant information for inclusion in the security reports.
The session includes a detailed walkthrough of how the combined use of RAG, fine-tuned LLMs, and Oracle Database 24c's vector search can automate the assimilation of vast amounts of data, transforming them into coherent, actionable security reports. Attendees gain insights into the practical application of these technologies in real-world security scenarios, illustrating how they can significantly reduce the time and effort required to produce security reports while simultaneously increasing their accuracy and depth of analysis.
This holistic approach streamlines the reporting process and enhances the overall security response by providing timely, data-driven insights into security incidents and breaches. The presentation aims to empower security professionals with the knowledge and tools necessary to leverage these cutting-edge technologies, ultimately strengthening their organization's security posture and incident response capabilities.ODSC Links:
• Get free access to more talks/trainings like this at Ai+ Training platform:
https://hubs.li/H0Zycsf0
• ODSC blog: https://opendatascience.com/
• Facebook: https://www.facebook.com/OPENDATASCI
• Twitter: https://twitter.com/_ODSC & @odsc
• LinkedIn: https://www.linkedin.com/company/open-data-science
• Slack Channel: https://hubs.li/Q02r9VZM0
• Code of conduct: https://odsc.com/code-of-conduct/