IISc CSA

An article titled “ SuperUVM: Empowering UVM with Observability”, authored by Pratheek B, Khushit Shah, and Arkaprava Basu, has been accepted to the 53rd IEEE/ACM International Symposium on Computer Architecture (ISCA 2026).

Polymage Labs, an IISc-incubated startup from CSA faculty Prof. Uday Kumar Reddy B, had an exhibit at the HiPEAC 2026 conference in Poland, which was covered by EE Times. Interview youtube.com/watch?v=8unA6t… News article eetimes.com/hipeac-2026-do… Polymage Labs polymagelabs.com

has been accepted for publication at the International Conference on Learning Representations (ICLR), 2026.

A paper titled "Ads that Stick: Near-Optimal Ad Optimization through Psychological Behavior Models" authored by Kailash Gopal Darmasubramanian, Akash Pareek, Arindam Khan, and Arpit Agarwal,

A paper titled "Approximation Schemes and Structural Barriers for the Two-Dimensional Knapsack Problem with Rotations," authored by Debajyoti Kar, Arindam Khan, and Andreas Wiese, has been accepted for publication at the 58th Annual ACM Symposium on Theory of Computing (STOC), 26

Paper titled "New Upper and Lower Bounds for Perfectly Secure MPC" by Ivan Damgård, Shravani Patil, Arpita Patra and Lawrence Roy has been accepted for publication at EUROCRYPT 2026

Accepted to #AISTATS2026: "Monotone and Conservative Policy Iteration Beyond the Tabular Case" by @EshwarSR , @GThoppe , A. Barua, Aditya Gopalan (@today_itself), and @DalalGal. This work extends PI guarantees to function approximation. arxiv.org/abs/2506.07134

Highly promising results here from Polymage's collaboration with Tenstorrent --- showing that the right compiler technology can be a leapfrog advance to the programming ecosystem for new chips! News: lnkd.in/gvpHq5rFA more technical version: lnkd.in/gqpjHxJ8

Polymage's PolyBlocks engine enables the rapid development of compilers for new AI chips, extracting close-to-peak performance from hardware, and realizing high-performance AI systems for vision, text, generative AI, and more.

where it received the Best Poster Award. doi.org/10.1051/epjcon…

Best Poster Award at Powders and Grains 2025. Paper titled 'Topological Signatures of Jamming in Granular Force Networks via Persistent Homology’ co-authored by Vishali S, Abrar Naseer, Vijay Natarajan, and Tejas G Murthy was presented at the conference Powders and Grains 2025,

The Walmart Centre of Tech Excellence at CSA proudly presents the 2026 edition of the Reinforcement Learning Workshop. This workshop aims to expose students and researchers in India to state-of-the-art Reinforcement Learning (RL) research. Register now: events.csa.iisc.ac.in/rlworkshop2026/

The registration deadline has been extended to December 25, 2025. We invite students, faculty, and researchers interested in reinforcement learning to register

Accepted to #AAAI2026: “Parameter-free Optimal Rates for Nonlinear Semi-Norm Contractions with Applications to Q-Learning,” by Ankur Naskar, Gugan Thoppe (@GThoppe), and Vijay Gupta. This work shows how to handle the non-monotonicity of semi-norms in #RL. arxiv.org/pdf/2508.05984

🔗 More info: events.csa.iisc.ac.in/csworkshop2025/ #SystemsResearch #IISc #ComputerSystems #ResearchCommunity

🎉 We’re hosting the Computer Systems Workshop 2025 at @IIScBangalore! 🧠 Exploring frontiers of systems research across industry & academia 🎤 Keynotes by Mark D. Hill (UW–Madison) & Manosiz Bhattacharyya (Nutanix) 🗓️ Oct 30–31 | 📍CSA Auditorium, IISc

🎉Congratulations to Prof. Arkaprava Basu on winning the "Rashtriya Vigyan Yuva Puraskar - Shanti Swarup Bhatnagar" for the year 2025. 🏆 He is the sole recipient this year in the Engineering Sciences category.

Does DQN Learn?” (arxiv.org/pdf/2205.13617) by Aditya Gopalan (@today_itself) and Gugan Thoppe (@GThoppe) has been accepted to the IEEE Transactions on Automatic Control. In practice, DQN often fails to identify good policies—this work provides a rigorous explanation for why.

Paper titled “Compatibility of Fairness and Nash Welfare under Subadditive Valuations’’ by Siddharth Barman and Mashbat Suzuki has been accepted for publication at SODA 2026: arxiv.org/abs/2407.12461