Version 1.0.0-alpha2 - Fix: Consistent scoring behavior in Evaluator and EvaluationSet Key changes: - Remove default scores fallback in Evaluator::calculateScore() - Now returns 0.0 for unknown choices (matching EvaluationSet::getScoreForChoice) - Update tests to reflect new behavior - All tests pass (121 unit tests, 131 kernel tests) This fixes an inconsistency where Evaluator::calculateScore() had a default scores fallback while EvaluationSet::getScoreForChoice() only returned 0.0 for unknown choices. Co-authored-by: openhands <openhands@all-hands.dev>