Repository navigation
Content audit: label the A* admissibility correction - #14
Merged
Merged
Conversation
The Part A report argued that both Manhattan and Euclidean heuristics were admissible. The paired study shows neither is on the hex grid, and the site said so without noting that the original report claimed the opposite. The landing page, the A* Lab findings and the README now say what the report claimed and that it was corrected in 2026. Content audit: every other landing, methods, agent-card, decision-record and tour figure was traced to the original code, notebooks, report or the reference-study data and matched. The original benchmark (144/160) was re-run with uv and reproduced byte for byte. Co-Authored-By: Claude Opus 5.5 <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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Content-consistency audit of https://cachex-ai.vercel.app against the original coursework.
Fixed (1 mismatch): the Part A report argued that both heuristics were admissible; the site said neither is, without saying the report claimed otherwise. The landing page, the A* Lab findings and the README now label this as corrected in 2026.
Checked and matching:
sample_output.txt,sample_output2.txt), dynamic depth thresholds 15/10/5% and the two-turn opening book (_4399/player.py), the six weights (utility/weights.json), Python 3.6 / NumPy / SciPy (environment.yml), Semester 1 2022 and the two-person team.uv run scripts/benchmark_agent.py, output byte-identical toagent-benchmark.json./tournament: Elo and win-rate table, head-to-heads, 34 of 4,602 deeper searches, alpha-beta 444/480 and 0.6% to 92%, textbook 17% to 45%, 980/931/107 A* boards, 11.3 pp (9.1 to 13.3), parity fixture counts (184, 40, 237, 179, 143, 5), referee limits (60 s, 100 MB, n 3 to 15), the beta-update bug, scripted-line counts (50/43/34, 124 of 127).Gates
pnpm lint,format:check,typecheck,test(989 passed) andbuildall pass locally.🤖 Generated with Claude Code