Docs / source reviewed
jevman
Let Jev play Pac-Man and compare decision models
How this project uses Jev
- Input
- Maze layout, Pac-Man and ghost positions, remaining pellets, game mode, power-ups and legal directions
- Jev decides
- Chooses a direction at junctions and answers whether to turn back in corridor-escape situations; probabilities and model latency appear on the board and in the decision log
- Code executes
- A browser simulator advances the game, with a recorded Jev replay, live play and model matchups; a rule-based move is used when an answer is not ready in time
Evidence and limitations
The AGPL-3.0 license, Jev API call, game-state summary and fallback logic were reviewed without running the site or model. The default view is an author-recorded game; live comparisons need Opper or TypeSafe credentials/credits. Scores and outcomes belong to this project’s game setup, not a general capability benchmark.
This project has not been run independently here. Author-reported results are not independently verified results.
Original sources
- https://github.com/opper-ai/jevman-benchmark/blob/1c8b77ac61bafa10956c601b1c9b339dcd277836/README.md ↗
- https://github.com/opper-ai/jevman-benchmark/blob/1c8b77ac61bafa10956c601b1c9b339dcd277836/LICENSE ↗
- https://github.com/opper-ai/jevman-benchmark/blob/1c8b77ac61bafa10956c601b1c9b339dcd277836/server/jev.ts ↗
- https://github.com/opper-ai/jevman-benchmark/blob/1c8b77ac61bafa10956c601b1c9b339dcd277836/src/brain.ts ↗
- https://github.com/opper-ai/jevman-benchmark/blob/1c8b77ac61bafa10956c601b1c9b339dcd277836/src/demo.ts ↗
- https://github.com/opper-ai/jevman-benchmark/blob/1c8b77ac61bafa10956c601b1c9b339dcd277836/server/routes.ts ↗