List your company

Reasoning and Classification on BBH (unseen)

98.8BBH Temporal SequencesRLHFlow-Llama3.1-8B-D
12.68835.04457.479.756May 23, 2023Sep 21, 2023Jan 21, 2024May 21, 2024Sep 20, 2024Jan 19, 2025May 21, 2025

Evaluation results

MethodLinks
RLHFlow-Llama3.1-8B-DK=72025.0598.8——————————————————————77.292.840
GPT-4Prompting=Direct, Setting=Zero-shot2023.059877.659.974.869.264.440.862.866.8669479.568.876.784.889.162.487.684.425.225.242.449.667.4——
FOVER-Llama3.1-8B-PRMK=72025.0597.2——————————————————————809646.8
FOVER-Qwen2.5-7B-PRMK=72025.0592.4——————————————————————70.79029.6
RLHFlow-Llama3.1-8B-MK=72025.0592——————————————————————7490.839.2
Qwen 2.5 7BK=72025.0591.2——————————————————————69.789.228.8
Llama 3.1 8BK=72025.0590.4——————————————————————749041.6
Qwen2.5-Math-PRM-7BK=72025.0586.4——————————————————————69.392.828.8
Qwen2.5-7B-Skywork-PRMK=72025.0584——————————————————————68.991.231.6
Qwen2.5-Math-7B-PRM800KK=72025.0582——————————————————————67.190.828.4
CODEXPrompting=Direct, Setting=Zero-shot2023.0577.688.463.663.667.252.43260.432.42652.884.850.466.467.675.26261.272.820.414.437.651.652.8——
CHATGPTPrompting=Direct, Setting=Zero-shot2023.0535.682.857.242.857.253.625.669.238.839.660.455.455.645.947.65640.85979.618.415.231.65648.9——
COT-T5-11BPrompting=Direct, Setting=Zero-shot, Parameters=11B2023.0528.859.260.25168.255.210.464.254.460.67552.86041.833.241.649.266.266.413.29.631.251.648——
FLAN-T5-11BPrompting=Direct, Setting=Zero-shot, Parameters=11B2023.0528.856.86154.867.255.221.270.853.66074.436.461.641.823.234.449.270.26015.21224.45047.05——
COT-T5-11BPrompting=CoT, Setting=Zero-shot, Parameters=11B2023.0527.465.660.45263.451.22265.248.252.455.444.65939.132.642.843.867.764.82018.441.85747.6——
FLAN-T5-11BPrompting=CoT, Setting=Zero-shot, Parameters=11B2023.0524.451.658.346.863.254.412.455.251.257.666.432.460.841.822.831.635.659.55619.613.228.852.843.32——
VICUNA-13BPrompting=Direct, Setting=Zero-shot, Parameters=13B2023.051640.842.21018.447.23.6444.81.216.843.446.415.11215.7228.148.49.25.623.241.223.3——