List your company

Factuality Correction on VELI5 1.0 (test)

36Precision (Pr)FACTCORRECTOR
-5.65.21626.8Jan 16, 2026

Evaluation results

MethodLinks
FACTCORRECTORbase model=gpt-oss-120b2026.01361930714
FACTCORRECTORbase model=mixtral-8x22b-instruct2026.0134924147
RACbase model=gpt-oss-120b2026.01341527613
LLM2base model=gpt-oss-120b2026.01332028513
LLM1base model=mixtral-8x22b-instruct2026.01321023127
LLM2base model=mixtral-8x22b-instruct2026.01321023137
LLM1base model=granite-4.0-h-small2026.01311123118
FACTCORRECTORbase model=granite-4.0-h-small2026.01291021117
FACTCORRECTORbase model=llama-3.3-70b-instruct2026.012771858
CRITICbase model=mixtral-8x22b-instruct2026.012681996
LLM2base model=llama-3.3-70b-instruct2026.012671858
LLM1base model=llama-3.3-70b-instruct2026.012561858
SFTbase model=llama-3.3-70b-instruct2026.01251119108
RACbase model=mixtral-8x22b-instruct2026.012481886
SFTbase model=granite-4.0-h-small2026.0122111897
LLM1base model=gpt-oss-120b2026.01211619-113
RACbase model=granite-4.0-h-small2026.012091667
LLM2base model=granite-4.0-h-small2026.0120101695
SFTbase model=mixtral-8x22b-instruct2026.0119101658
CRITICbase model=llama-3.3-70b-instruct2026.011961335
CRITICbase model=granite-4.0-h-small2026.011741256
RACbase model=llama-3.3-70b-instruct2026.01132925
SFTbase model=gpt-oss-120b2026.0188872
CRITICbase model=gpt-oss-120b2026.01-41-2-1-3