Source record

Published by arXiv cs.MA (T1 source). The original is at https://arxiv.org/abs/2607.20501.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

MKEvolve is a multi-agent framework that iteratively co-evolves a modular decomposition of PyTorch modules with LLM-generated Triton kernels, refining both the structure and each subkernel via beam search. On KernelBench L2/L3 it improves correctness and speedup over end-to-end synthesis while cutting LLM token usage by up to 35%.

Reports concrete benchmark deltas (35% token reduction, correctness and speedup gains) on KernelBench L2/L3, useful as a data point on multi-agent code generation tradeoffs.

Cited by

No citations on record.