List your company

Explicit-Implicit Reasoning on MixRea

76.5IIAGemini 2.5 Pro
54.03659.86865.771.532May 19, 2026

Evaluation results

MethodLinks
Gemini 2.5 ProAlignment=chat or instruct2026.0576.591.49.4
LLaMA 3.1-405BAlignment=chat or instruct2026.0571.191.112.4
Deepseek-R1Alignment=chat or instruct2026.0570.992.49.5
Claude 3.7 SonnetAlignment=chat or instruct2026.0570.192.112.7
LLaMA 3.1-70BAlignment=chat or instruct2026.0567.690.513.6
Gemini 2.5 FlashAlignment=chat or instruct2026.0567.192.515.1
GPT-4oAlignment=chat or instruct2026.0566.691.513.2
Qwen 2.5-32BAlignment=chat or instruct2026.0566.491.714.6
Deepseek-V3Alignment=chat or instruct2026.0565.492.514
LLaMA 3-70BAlignment=chat or instruct2026.0565.490.314.6
Gemma 3-27BAlignment=chat or instruct2026.0565.287.616.2
Gemma 2-27BAlignment=chat or instruct2026.0565.187.214.3
Qwen-MaxAlignment=chat or instruct2026.056591.615.4
Qwen 2-7BAlignment=chat or instruct2026.0564.285.114.3
Qwen 2.5-72BAlignment=chat or instruct2026.0563.987.916
Qwen 2.5-7BAlignment=chat or instruct2026.0562.684.115
Gemma 3-12BAlignment=chat or instruct2026.0561.886.618.2
Qwen 2-72BAlignment=chat or instruct2026.0561.391.617.7
GPT-4o-miniAlignment=chat or instruct2026.0557.586.620.3
LLaMA 3-8BAlignment=chat or instruct2026.0557.581.718.3
LLaMA 3.1-8BAlignment=chat or instruct2026.0554.983.320.3