Search results

MNIST with rolled kernels takes 7 minutes per inference

Status: Open.

#2975 In google/heir;

· j2kun opened on May 20, 2026

clean-up: consider removing domain schedule from tensor_ext.assign_layout

Status: Open.

#2973 In google/heir;

· asraa opened on May 20, 2026

follow-up: handle zero/splat optimizations for layout attr lowering

Status: Open.

#2965 In google/heir;

· asraa opened on May 19, 2026

experimental-disable-loop-unroll is incompatible with split-preprocessing

Status: Open.

#2960 In google/heir;

· j2kun opened on May 18, 2026

Check permutation map for convolution operations

Status: Open.

#2957 In google/heir;

· mdgrs opened on May 18, 2026

Implement Goldschmidt iteration as special case for math.div %cst_1, %x (inverse)

Status: Open.

#2936 In google/heir;

· AlexanderViand opened on May 11, 2026

Special-case math.exp approximation

Status: Open.

#2935 In google/heir;

· AlexanderViand opened on May 11, 2026

Add codegen for Jaxite CKKS backend

Status: Open.

#2932 In google/heir;

· j2kun opened on May 8, 2026

Complexification Optimizations

research synthesisReading papers to figure out which ideas can be incorporatedReading papers to figure out which ideas can be incorporated

Status: Open.

#2930 In google/heir;

· lawrencekhlim opened on May 8, 2026

Improve torch operator coverage (as exported by torch-mlir)

torch support

Status: Open.

#2923 In google/heir;

· j2kun opened on May 6, 2026

Use centered representatives for basis conversion

Status: Open.

#2922 In google/heir;

· crockeea opened on May 6, 2026

RNS::ExtractSliceOp is unused

Status: Open.

#2921 In google/heir;

· crockeea opened on May 5, 2026