[FM-India]IARCS Verification Seminar Series -- Talk by Kishor Jothimurugan on June 20 at 1900 hrs IST

Dear all, The next talk in the IARCS Verification Seminar Series will be given by Kishor Jothimurugan, an incoming Quantitative Researcher at Two Sigma. The talk is scheduled on Tuesday, June 20, at 1900 hrs IST (add to Google calendar <https://calendar.google.com/calendar/event?action=TEMPLATE&tmeid=NjViM244cWxkN3FyM2xudHRwcGM1bnNnb2MgdnNzLmlhcmNzQG0&tmsrc=vss.iarcs%40gmail.com> ). The details of the talk can be found on our webpage ( https://fmindia.cmi.ac.in/vss/), and also appended to the body of this email. The Verification Seminar Series, an initiative by the Indian Association for Research in Computing Science (IARCS), is a monthly, online talk-series, broadly in the area of Formal Methods and Programming Languages, with applications in Verification and Synthesis. The aim of this talk-series is to provide a platform for Formal Methods researchers to interact regularly. In addition, we hope that it will make it easier for researchers to explore newer problems/areas and collaborate on them, and for younger researchers to start working in these areas. All are welcome to join. Best regards, Akash, Deepak, Madhukar, Srivathsan ============================================================= Title: Specification-Guided Reinforcement Learning Meeting Link: https://us02web.zoom.us/j/89164094870?pwd=eUFNRWp0bHYxRVpwVVNoVUdHU0djQT09 (Meeting ID: 891 6409 4870, Passcode: 082194) Abstract: Recent advances in Reinforcement Learning (RL) have enabled data-driven controller design for autonomous systems such as robotic arms and self-driving cars. Applying RL to such a system typically involves encoding the objective using a reward function (mapping transitions of the system to real values) and then training a neural network controller (from simulations of the system) to maximize the expected reward. However, many challenges arise when we try to train controllers to perform complex long-horizon tasks---e.g., navigating a car along a complex track with multiple turns. Firstly, it is quite challenging to manually define well-shaped reward functions for such tasks. It is much more natural to use a high-level specification language such as Linear Temporal Logic (LTL) to specify these tasks. Secondly, existing algorithms for learning controllers from logical specifications do not scale well to complex tasks due to a number of reasons including the use of sparse rewards and lack of compositionality. Furthermore, existing algorithms for verifying neural network policies (trained using RL) cannot be easily applied to verify policies for complex long-horizon tasks due to large approximation errors. In this talk, I will present my work on using logical specifications to specify RL tasks. First, I'll talk about algorithms for learning control policies from such specifications. Then, I'll show how we can use logical task decompositions to scale verification to long-horizons. Bio: Kishor Jothimurugan is an incoming Quantitative Researcher at Two Sigma. He earned his PhD in Computer and Information Science from the University of Pennsylvania, where he was advised by Prof. Rajeev Alur. His research interests lie at the intersection of Formal Methods and Machine Learning. In particular, he is interested in applying formal methods to improve applicability and reliability of reinforcement learning, verifying systems with neural network components and using neurosymbolic approaches to improve program synthesis and analysis.