docs: benchmark findings v1 (nullcheck baseline-vs-JIT)#12
Merged
Conversation
Records the first real baseline-vs-JIT measurement and its conclusion so it's preserved beyond chat: on the trivial `nullcheck` shape a project- local skill saves ~354 tokens/use (~0.15%) at iso-accuracy, so break-even is thousands of invocations — not worth compiling for this shape. Captures method (fresh-per-arm, T2S from API usage, iso-accuracy + dual gate), the numbers, interpretation (baseline dominated by fixed context overhead), the hypothesis for where JIT should pay off (exploration-heavy shapes like migrate-N), how to reproduce, and limitations/next steps. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Records the first real baseline-vs-JIT measurement from
aj benchand its conclusion, so the finding is preserved in the repo (not just chat).Headline: on the trivial
nullcheckshape, a project-local skill saves ~354 tokens/use (~0.15%) at 100%/100% iso-accuracy → break-even is thousands of invocations, i.e. not worth compiling for this shape. Baseline T2S (~236k) is dominated by fixed context/system-prompt overhead, so shortcutting a 2-line edit barely helps.Doc:
docs/superpowers/specs/2026-07-06-benchmark-findings.md. Covers method, the numbers, interpretation, the hypothesis for where JIT should pay off (exploration-heavy shapes likemigrate-N), reproduction, and limitations/next steps.Docs-only; no code change.
🤖 Generated with Claude Code