Implement test to quantify the impact of different features on gradient approximation

Problem KernelCI Debos rootfs images currently use a generic shell prompt: / # Some newer images may instead show: ~ # KernelCI lists these strings as expected login prompts in its LAVA job templates. LAVA warns that these prompts are not distinctive and may accidentally match…

Description With the PyTorch backend now supporting jit (see issue #4), it is crucial to establish a comprehensive benchmark suite. This suite will be used to evaluate and compare the performance of the PyTorch backend agaisnt TensorFlow, and JAX backends across different…

Problem Maintainer rule (2026-11-04): "we do not cancel ci in the same prs, the rule states, 2 prs may run ci at the same time the amount of cis in those are not limited" and "so we park the whole pr not the ci jobs, that is one source of slowness". The CI queue gate (a local…

Summary Opening a repo in Codebase Analytics starts a fresh scan and discards what was there before. There is no way to see whether results improved or regressed between runs, because the previous run no longer exists by the time the new one finishes. Audited before filing — this…