Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn about tree PCPs, a novel extension of probabilistically checkable proofs that evolve dynamically as computation progresses in this research talk. Discover how tree PCPs allow encoding computations so they can be quickly verified by reading only a few symbols, with proofs that grow incrementally over time by appending short strings to previous proofs. Explore the construction of tree PCPs for non-deterministic space-s computation, where proofs grow by only poly(s,log(t)) bits at each time step t, while maintaining efficient verification with poly(s) * t^epsilon queries for any constant epsilon > 0. Understand how tree PCPs serve as information-theoretic analogs to incrementally verifiable computation and examine their compilation in the random oracle model to achieve variants of IVC where provers can make limited queries to evolving states. Gain insights into this collaborative research between Tamer Mour, Alon Rosen, and Ron Rothblum that bridges theoretical computer science concepts of proof systems with practical applications for ongoing computational verification.
