Jeremy Kun
Jeremy Kun
Cf. https://github.com/google/heir/issues/396 This WIP PR is intended to explore running Polygeist on a variety of C++ input programs, and attempting to feed those as input to HEIR passes, so we...
It's currently ```mlir secret.generic ins(%arg0, %arg1 : secret, secret) { ^bb0(%arg0: ty0, %arg1: ty1): ...ops... } ``` The discrepancy between how the `ins` part is parsed/printed vs the basic block...
Current dep is v1.1.2, which has some known issues with pke tests for scheme switching (see https://github.com/openfheorg/openfhe-development/pull/659). To get an end-to-end build system working inside Google and externally, we'll need...
Re-enable the tests of the ML workflow once HEIR is in a place to execute it without it taking ~hours to run.
In-progress work adding python bindings to HEIR
Specific to NN quantization, but may have some useful generic ideas > an intra-coefficient packing algorithm and a quantization-aware tiling algorithm https://arxiv.org/abs/2401.15970 Abstract: > Secure two-party computation with homomorphic encryption...
After [additional reading](https://discourse.llvm.org/t/rfc-polygeist-llvm-incubator-proposal/60890), Polygeist seems like it is poised to be the primary bridge between ClangAST and MLIR, while ClangIR will continue to lower directly to LLVM. This implies we...
https://eprint.iacr.org/2023/1642 A New Perspective on Key Switching for BGV-like Schemes > Furthermore, we revisit an idea by Gentry, Halevi, and Smart [19] switching primes in and out of the ciphertext...
Tests to add: - https://github.com/google/heir/pull/203/files/2b456dc526c2c5a3b95fdfc4777c61b6cc582d24#r1378833393