List your company

Hallucination Detection on SQuAD (test)

83.8AUROCrHALLUGUARD
58.871265.343171.81578.2869Jan 26, 2026

Evaluation results

MethodLinks
HALLUGUARDBackbone=Llama2-70B2026.0183.881.7770.4673.24
HALLUGUARDBackbone=Llama2-13B2026.0181.4578.3964.3965.07
InsideBackbone=Llama2-70B2026.0181.2475.0969.4862.4
HALLUGUARDBackbone=Llama2-7B2026.0181.0577.1671.1864.38
HALLUGUARDBackbone=Llama-3-8B2026.0179.5678.2967.9763.27
InsideBackbone=Llama-3-8B2026.0176.1372.4465.6262.94
InsideBackbone=Llama2-13B2026.0174.6874.8161.0159.51
PerplexityBackbone=Llama2-70B2026.0174.2370.8862.2458.05
SelfCheckGPTBackbone=Llama-3-8B2026.0173.9972.1565.2654.02
InsideBackbone=Llama2-7B2026.0173.6375.7465.2259.11
MINDBackbone=Llama2-70B2026.0173.4671.7157.7656.77
SelfCheckGPTBackbone=Llama2-70B2026.0173.0770.4956.5954.65
LN-EntropyBackbone=Llama2-70B2026.0172.4468.9156.7752.63
Semantic Ent.Backbone=Llama2-70B2026.0172.0168.5156.4950.9
P(true)Backbone=Llama2-13B2026.0171.5668.457.5145.66
RACEBackbone=Llama2-7B2026.0171.3569.2359.1854.73
FActScoreBackbone=Llama-3-8B2026.0171.269.4561.9254.91
Semantic Ent.Backbone=Llama2-13B2026.0170.5867.2253.3152.94
PerplexityBackbone=Llama-3-8B2026.0170.5167.5155.7152.68
FActScoreBackbone=Llama2-7B2026.0170.3268.6358.1353.01
Lexical Sim.Backbone=Llama2-7B2026.0170.3169.0853.9753.31
PerplexityBackbone=Llama2-13B2026.0170.1969.2260.3354.82
EnergyBackbone=Llama2-70B2026.0169.0166.1958.4449.82
Lexical Sim.Backbone=Llama2-70B2026.0168.9567.9160.5256.56
MINDBackbone=Llama2-13B2026.0168.6466.9554.9252.49
Lexical Sim.Backbone=Llama2-13B2026.0168.5367.4250.7354.12
FActScoreBackbone=Llama2-70B2026.0168.3365.2656.9348.46
SelfCheckGPTBackbone=Llama2-7B2026.0168.2667.0960.0657.31
RACEBackbone=Llama-3-8B2026.0168.1766.0254.6553.06
P(true)Backbone=Llama-3-8B2026.0167.4263.9455.3547.52
P(true)Backbone=Llama2-70B2026.0166.8162.7157.4346.85
FActScoreBackbone=Llama2-13B2026.0166.6563.256.4153.42
Semantic Ent.Backbone=Llama-3-8B2026.0166.5262.6252.3752.7
Lexical Sim.Backbone=Llama-3-8B2026.0166.4363.5653.1950.96
LN-EntropyBackbone=Llama2-7B2026.0165.9664.2253.4352.84
SelfCheckGPTBackbone=Llama2-13B2026.0165.4761.6553.1249.89
MINDBackbone=Llama2-7B2026.0164.5761.1152.3953.13
EnergyBackbone=Llama-3-8B2026.0164.4161.0256.1746.21
RACEBackbone=Llama2-13B2026.0164.1960.4547.5345.66
RACEBackbone=Llama2-70B2026.0164.0562.3954.3850.07
PerplexityBackbone=Llama2-7B2026.0163.9361.7746.9748.2
LN-EntropyBackbone=Llama-3-8B2026.0163.760.446.1942.85
P(true)Backbone=Llama2-7B2026.0162.5561.0946.8452.32
MINDBackbone=Llama-3-8B2026.0162.2959.5844.4948.61
LN-EntropyBackbone=Llama2-13B2026.0161.6659.1649.0546.27
EnergyBackbone=Llama2-13B2026.0161.0259.7348.2642.08
Semantic Ent.Backbone=Llama2-7B2026.0160.2957.7343.6348.83
EnergyBackbone=Llama2-7B2026.0159.8356.1146.1943.18