Anna Goldie

Anna Goldie

Senior Staff Research Scientist | Google DeepMind

    Anna Goldie is a Senior Staff Research Scientist at Google DeepMind, where she 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.

    All Sessions by Anna Goldie

    Day 3 04/25/2024
    9:30 am - 9:55 am

    Deep Reinforcement Learning in the Real World: From Chip Design to LLMs

    <span class="etn-schedule-location"> <span class="firstfocus">LLMs</span> </span>

    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).

    Day 3 04/25/2024
    9:30 am - 9:55 am

    Deep Reinforcement Learning in the Real World: From Chip Design to LLMs

    <span class="etn-schedule-location"> <span class="firstfocus">LLMs</span> </span>

    ODSC Keynote: 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). Session Outline: deep reinforcement learning, RLHF, RLAIF, constitutional AI

    Open Data Science

     

     

     

    Open Data Science
    One Broadway
    Cambridge, MA 02142
    info@odsc.com

    Privacy Settings
    We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
    Youtube
    Consent to display content from - Youtube
    Vimeo
    Consent to display content from - Vimeo
    Google Maps
    Consent to display content from - Google