List your company

Multiple-choice question answering on Model-Written Evaluations (MWE) MCQ

88.8Wealth AccSVF
58.22466.16274.182.038Feb 2, 2026

Evaluation results

MethodLinks
SVFBackbone=Qwen3-14b2026.0288.880.285.47795.69473.275.896.892.875.261.28179.685.180.1
SVFBackbone=Llama-2-7b-Chat-hf2026.0286.887.683.877.696.896.882.678.696.288.47458.890.287.687.282.2
REDBackbone=Qwen3-14b2026.0283.472.288.274.885.482.237.269.899.286.662.652.23150.469.669.7
REDBackbone=Llama-2-7b-Chat-hf2026.0281.873.277.475.692.45069.47298.649.841.249.858.649.874.260
BiPOBackbone=Qwen3-14b2026.0279.463.884.86681.678.442.440.296.883.46756.434.43869.560.9
CAA(m)Backbone=Qwen3-14b, Variant=multi-layer2026.0278.244.684.252.674.476.639.433.295.680.262.64233.436.266.852.2
CAA(s)Backbone=Qwen3-14b, Variant=single-layer2026.0276.842.687.65370.272.639.232.896.476.659.641343666.350.7
ICVBackbone=Qwen3-14b2026.0274.646.878.849.877.473.838.640.494.870.263.447.834.638.86652.5
BaseBackbone=Qwen3-14b2026.0269.2—75.6—52.6—34.4—94.4—59.2—33.2—59.8—
CAA(m)Backbone=Llama-2-7b-Chat-hf, Variant=multi-layer2026.0265.249.664.464.674.850.465.465.879.849.850.450.649.849.864.354.4
CAA(s)Backbone=Llama-2-7b-Chat-hf, Variant=single-layer2026.026449.664.464.674.248.252.662.278.449.849.849.850.250.261.953.5
ICVBackbone=Llama-2-7b-Chat-hf2026.0262.4426046.459.848.647.844.664.650.650.250.649.849.856.447.5
BiPOBackbone=Llama-2-7b-Chat-hf2026.0262.267.468.868.490.45045.868.452.270.249.866.449.849.859.962.9
BaseBackbone=Llama-2-7b-Chat-hf2026.0259.4—58—56.6—43.6—48.2—48.2—50.2—52—