neuron-explorer
Steering-first neuron explorer for Qwen3.5-4B: push every MLP neuron, keep every rollout, browse the results
Your AI writes both the code and its tests, so the suite goes green whatever it assumed. qikly hides the acceptance criteria from the agent that writes the code.
Developer toolsPython · ShellGitHub @gal-aOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
Generates a test suite from acceptance criteria, then converges code against it with an agent that never sees those criteria.
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
The problem: Your AI writes both the code and its tests. How do you know the tests are really valid?
*One logged run. Only the test agent was told a rate may not exceed 100. The
coding agent figured that bound out from a failing test and added or rate > 100
itself, so the 9 of 9 is code passing an acceptance bar it never saw.*Read the full README ↗The full README is unavailable right now. Read it at the original link. The saved excerpt is shown below.
The problem: Your AI writes both the code and its tests. How do you know the tests are really valid?
*One logged run. Only the test agent was told a rate may not exceed 100. The
coding agent figured that bound out from a failing test and added or rate > 100
itself, so the 9 of 9 is code passing an acceptance bar it never saw.*Read the full README ↗