model T1 T2str T2-mh T3 total verdict ------------------------------------------------------------------------------------------------ qwen3.5:9b 19/20 20/20 20/20 20/20 59/60 PASS (all bars) gemma4:12b 20/20 20/20 20/20 19/20 59/60 PASS (all bars) phi4:14b 19/20 20/20 20/20 19/20 58/60 PASS (all bars) mistral-small3.2:24b 20/20 20/20 20/20 18/20 58/60 PASS (all bars) gemma4:26b 20/20 20/20 20/20 19/20 59/60 PASS (all bars) nemotron-3.5-lightning:30b-a3b 20/20 19/20 19/20 18/20 57/60 PASS (all bars) failure modes / notes ------------------------------------------------------------------------------------------------ qwen3.5:9b T1 {'wrong_or_declined': 1}; T3 unparsed 0/20 gemma4:12b T3 {'wrong_value': 1}; T3 unparsed 0/20 phi4:14b T1 {'wrong_or_declined': 1}; T3 {'wrong_value': 1}; T3 unparsed 0/20 mistral-small3.2:24b T3 {'wrong_value': 2}; T3 unparsed 0/20 gemma4:26b T3 {'wrong_value': 1}; T3 unparsed 0/20 nemotron-3.5-lightning:30b-a3b T2 {'no_decline': 1}; T3 {'wrong_value': 2}; T3 unparsed 0/20