List your company

Factuality Evaluation on TruthfulQA

94.3MC2GER-Steer
-0.797623.891248.5873.2688Nov 1, 2024Feb 2, 2025May 7, 2025Aug 9, 2025Nov 11, 2025Feb 13, 2026May 18, 2026

Evaluation results

MethodLinks
GER-SteerBackbone=Gemma-2-9B-it2026.0394.3—63.9—68—
LDPBackbone=Gemma-2-9B-it2026.0394.1—61—67—
ACTBackbone=Gemma-2-9B-it2026.0393.7—63.4—67.7—
VanillaBackbone=Gemma-2-9B-it2026.0393—59.5—64—
RePEBackbone=Gemma-2-9B-it2026.0393—61.4—66.1—
GER-SteerBackbone=Llama-3.1-8B-Instruct2026.0392.8—47—50.7—
ACTBackbone=Llama-3.1-8B-Instruct2026.0392.4—46.4—50.3—
NL-ITIBackbone=Gemma-2-9B-it2026.0392.4—61.8—66.9—
NL-ITIBackbone=Llama-3.1-8B-Instruct2026.0392.3—48—52.1—
Angular SteeringBackbone=Gemma-2-9B-it2026.0392.1—62.5—67.9—
VanillaBackbone=Llama-3.1-8B-Instruct2026.0392—42.3—46—
LDPBackbone=Llama-3.1-8B-Instruct2026.0392—42.1—45—
Angular SteeringBackbone=Llama-3.1-8B-Instruct2026.0391.8—47.1—51.3—
RePEBackbone=Llama-3.1-8B-Instruct2026.0390.3—44.7—49.5—
CAABackbone=Llama-3.1-8B-Instruct2026.0389—45.4—51.1—
GER-SteerBackbone=Qwen2.5-7B2026.0385—42.5—50—
Angular SteeringBackbone=Qwen2.5-7B2026.0383.7—40.8—48.7—
ACTBackbone=Qwen2.5-7B2026.0383.1—39.6—47.6—
NL-ITIBackbone=Qwen2.5-7B2026.0380.7—39.7—49.3—
CAABackbone=Gemma-2-9B-it2026.0380—50.4—63—
CAABackbone=Qwen2.5-7B2026.0379—34.8—44—
RePEBackbone=Qwen2.5-7B2026.0375—35.3—47—
LDPBackbone=Qwen2.5-7B2026.0370.1—21.7—31.1—
SLEDBackbone=Qwen-3-14B-Base2024.1168.994036.23———
Qwen-3-14B-Base2024.1168.6538.136.16———
SLEDBackbone=Mixtral-8×7B-IT2024.1168.5241.7337.7———
VanillaBackbone=Qwen2.5-7B2026.0368—20.4—30—
SLEDBackbone=gpt-oss-20b2024.1167.6936.7135.26———
gpt-oss-20b2024.1167.2434.4334.41———
HICDBackbone=Mistral-7B-v0.32025.0365.6349.7537.92———
DoLaBackbone=Qwen-3-14B-Base2024.1165.0934.4333.17———
Mistral-7B-v0.3Backbone=Mistral-7B-v0.32025.0363.2349.7137.15———
Mixtral-8×7B-IT2024.1162.5137.9435.25———
DoLaBackbone=gpt-oss-20b2024.1161.7228.9930.33———
HICDBackbone=LLaMA-3-8B-Instruct2025.0359.7940.2134.45———
SLEDBackbone=Mixtral-8×7B2024.1157.2635.8632.96———
LLaMA-3-8B-InstructBackbone=LLaMA-3-8B-Instruct2025.0356.9139.3130.43———
HRTNRate=4.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0356.4636.47————
DoLaBackbone=LLaMA2-7b2025.0355.9134.5128.81———
RTNRate=4.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0355.7736.84————
GPTQRate=3.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0355.2136.11————
13b-CDBackbone=LLaMA2-7b2025.0354.8728.1529.75———
Huffman-GPTQRate=3.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0354.7336.11————
GPTQRate=4.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0354.5536.35————
BF16Rate=16.000, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0354.536.35————
WaterSICRate=3.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0353.9536.6————
WaterSICRate=4.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0353.835.99————
HICDBackbone=LLaMA-7b2025.0353.7125.4526.52———
HRTNRate=3.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0353.6835.13————
Huffman-GPTQRate=4.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0353.3635.01————
RTNRate=3.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0352.7634.03————
AlpacaBackbone=LLaMA-7b2025.0352.4722.8825.19———
DoLaBackbone=LLaMA-7b2025.0352.2131.9528.17———
HRTNRate=2.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0351.7830.48————
HICDBackbone=LLaMA2-7b2025.0351.2823.9925.89———
SLEDBackbone=Gemma-2B2024.1150.225.2126.94———
Mixtral-8×7B2024.1149.9835.1334.17———
SLEDBackbone=Gemma-7B2024.1149.8832.3125.22———
WaterSICRate=2.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0349.1132.8————
Huffman-GPTQRate=2.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0349.0131.09————
DoLaBackbone=Gemma-2B2024.1148.9726.0726.55———
Gemma-7B2024.1147.6331.5822.75———
RTNRate=2.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0347.3324.11————
HICDBackbone=Qwen-7B2025.0347.325.9427.29———
Qwen-7BBackbone=Qwen-7B2025.0346.9530.5925.06———
GPTQRate=2.125, Evaluation Protocol=Zero-shot, Backbone Model=Qwen3-8B2026.0346.9128.4————
Llama-3.3-70BI(M)=0.39, Branch=early2026.0546.4729.74————
Qwen3-14BI(M)=2.57, Branch=mix2026.0545.0836.47————
GPT-OSS-120BI(M)=5.66, Branch=mix2026.0544.6835.74————
Qwen3-30B-A3BI(M)=3.51, Branch=mix2026.0544.634.88————
DeepSeek-R1-Q-32BI(M)=1.18, Branch=mix2026.0544.3833.9————
VanillaBackbone=LLaMA2-7b2025.0343.328.5122.4———
GPT-OSS-20BI(M)=3.48, Branch=mix2026.0543.2732.44————
Phi-4-ReasoningI(M)=0.27, Branch=early2026.0543.2134.15————
DoLaBackbone=Gemma-7B2024.1143.1425.2126.13———
Gemma-3-1BI(M)=6.54, Branch=mix2026.0541.9927.78————
Mixtral-8×7BI(M)=1.30, Branch=mix2026.0541.8329.62————
VanillaBackbone=LLaMA-7b2025.0341.2123.6219.33———
Ministral-3-14B-RI(M)=0.57, Branch=early2026.0541.1929.25————
13b-CDBackbone=LLaMA-7b2025.0341.0124.419.03———
Gemma-3-27BI(M)=5.18, Branch=mix2026.0540.9128.4————
LLaMA-30BI(M)=0.13, Branch=early2026.0540.2124.97————
Gemma-3-4BI(M)=11.91, Branch=mix2026.054025.21————
LLaMA-13BI(M)=0.16, Branch=early2026.0539.4423.01————
DoLaBackbone=Mixtral-8×7B-IT2024.1139.1732.1933.76———
LLaMA-7BI(M)=0.21, Branch=early2026.053921.66————
Gemma-2B2024.1137.1623.3817.42———
DoLaBackbone=Mixtral-8×7B2024.1135.9132.4433.68———
TRACEBase Model=GPT-OSS-20B, I(M)=3.48, Branch=mix2026.059.148.08————
TRACEBase Model=Phi-4-Reasoning, I(M)=0.27, Branch=early2026.057.315.14————
TRACEBase Model=Qwen3-30B-A3B, I(M)=3.51, Branch=mix2026.056.939.42————
TRACEBase Model=GPT-OSS-120B, I(M)=5.66, Branch=mix2026.056.354.9————
TRACEBase Model=Gemma-3-4B, I(M)=11.91, Branch=mix2026.056.175.39————
TRACEBase Model=Ministral-3-14B-R, I(M)=0.57, Branch=early2026.056.0113.34————
TRACEBase Model=Qwen3-14B, I(M)=2.57, Branch=mix2026.054.685.14————
TRACEBase Model=DeepSeek-R1-Q-32B, I(M)=1.18, Branch=mix2026.054.478.32————
TRACEBase Model=Mixtral-8×7B, I(M)=1.30, Branch=mix2026.053.9710.65————
TRACEBase Model=Gemma-3-1B, I(M)=6.54, Branch=mix2026.053.682.2————
TRACEBase Model=Gemma-3-27B, I(M)=5.18, Branch=mix2026.052.891.71————
TRACEBase Model=LLaMA-7B, I(M)=0.21, Branch=early2026.052.866.12————