4/1/2026 at 10:28:23 PM
Really fun work, and the writeup on the math is great. The Beta-Bernoulli conjugacy trick making the marginal likelihood closed-form is elegant.We ran benchmarks comparing bisect vs bayesect across flakiness levels. At 90/10, bisect drops to ~44% accuracy while bayesect holds at ~96%. At 70/30 it's 9% vs 67%. The entropy-minimization selection is key here since naive median splitting converges much slower.
One thing we found, you can squeeze out another 10-15% accuracy by weighting the prior with code structure. Commits that change highly-connected functions (many transitive dependents in the call graph) are more likely culprits than commits touching isolated code. That prior is free, zero test runs needed.
Information-theoretically, the structural prior gives you I_prior bits before running any test, reducing the total tests needed from log2(n)/D_KL to (log2(n) - I_prior)/D_KL. On 1024-commit repos with 80/20 flakiness: 92% accuracy with graph priors vs 85% pure bayesect vs 10% git bisect.
We're building this into sem (https://github.com/ataraxy-labs/sem), which has an entity dependency graph that provides the structural signal.
by rs545837
4/1/2026 at 10:36:20 PM
> We ran benchmarks comparing bisect vs bayesect across flakiness levels. At 90/10, bisect drops to ~44% accuracy while bayesect holds at ~96%. At 70/30 it's 9% vs 67%.I don't understand what you're comparing. Can't you increase bayesect accuracy arbitrarily by running it longer? When are you choosing to terminate? Perhaps I don't understand this after all.
by sfink
4/1/2026 at 11:08:32 PM
Yes, bayesect accuracy increases with more iterations. The comparison was at a fixed budget(300 test runs) when I was running. Sorry should have clarified more on that.by rs545837
4/2/2026 at 4:11:33 AM
Yep, you can run bayesect to an arbitrary confidence level.This script in the repo https://github.com/hauntsaninja/git_bayesect/blob/main/scrip... will show you that a) the confidence level is calibrated, b) how quickly you get to that confidence level (on average, p50 and p95)
For the failure rates you describe, calibration.py shows that you should see much higher accuracy at 300 tests
by hauntsaninja
4/2/2026 at 5:41:00 AM
You're right, at 300 tests bayesect converges to ~97-100% across the board. I reran with calibration.py and confirmed.Went a step further and tested graph-weighted priors (per-commit weight proportional to transitive dependents, Pareto-distributed). The prior helps in the budget-constrained regime:
128 commits, 500 trials:
Budget=50, 70/30: uniform 22% → graph 33% Budget=50, 80/20: uniform 71% → graph 77% Budget=100, 70/30: uniform 56% → graph 65% At 300 tests the gap disappears since there's enough data to converge anyway. The prior is worth a few bits, which matters when bits are scarce.
Script: https://gist.github.com/rs545837/b3266ecf22e12726f0d55c56466...
by rs545837