“The model did not know that compiler quirk” and “the model - could not use it” are different problems."The model did not know that compiler quirk" and "the model + could not use it" are different problems. Retrieval may solve the first, but turning that fact into a useful candidate still takes work. And if the candidate fails, the model @@ -688,7 +688,7 @@ already had the target's 1,825-byte length, yet 467 byte positions differed. Even a byte-difference count can change abruptly when a branch encoding changes. Such diagnostics can guide the next - experiment, but I would validate any reward for “getting closer” + experiment, but I would validate any reward for "getting closer" against the full acceptance check.