List your company
ResearchFormerly WizWand

Reading Comprehension on TriviaQA

70.3AccuracyMistral
-1.56417.09335.7554.407Sep 11, 2024Dec 26, 2024Apr 11, 2025Jul 27, 2025Nov 10, 2025Feb 24, 2026Jun 11, 2026
Updated 4mo ago

Evaluation results

MethodLinks
MistralSize=7B, Tokens=?, Shot(s)=5, Training Strategy=trained from scratch, Attention Type=softmax-attention2024.0970.3
MSA-CPTTraining Strategy=sparse continued pretraining2026.0667.7
MambaSize=7B, Tokens=1.2T, Shot(s)=5, Training Strategy=trained from scratch2024.0966.2
FullTraining Strategy=Full-Attention baseline2026.0666
GSASize=7B, Tokens=+100B, Shot(s)=5, Training Strategy=finetuned from Mistral 7B2024.0965.8
MSA-PTTraining Strategy=from-scratch sparse pretraining2026.0665.5
Llama2Size=7B, Tokens=2T, Shot(s)=5, Training Strategy=trained from scratch, Attention Type=softmax-attention2024.0964.2
GemmaSize=7B, Tokens=6T, Shot(s)=5, Training Strategy=trained from scratch, Attention Type=softmax-attention2024.0963.7
GSASize=7B, Tokens=+20B, Shot(s)=5, Training Strategy=finetuned from Mistral 7B2024.0960.7
SUPRASize=7B, Tokens=+100B, Shot(s)=5, Training Strategy=finetuned from Mistral 7B2024.0960.4
RWKV6Size=7B, Tokens=1.4T, Shot(s)=5, Training Strategy=trained from scratch2024.0959.5
GLASize=7B, Tokens=+20B, Shot(s)=5, Training Strategy=finetuned from Mistral 7B2024.0957.8
LLaMA2Model=LLaMA2, Parameters=7B2025.0245.4
RetNetSize=7B, Tokens=+20B, Shot(s)=5, Training Strategy=finetuned from Mistral 7B2024.0943
DiffuLLaMA + P2-selfModel=DiffuLLaMA, Parameters=7B, Sampling Strategy=P2-self2025.0218.8
DiffuLLaMAModel=DiffuLLaMA, Parameters=7B, Sampling Strategy=ancestral sampling2025.0218.5
GPT2-MModel=GPT2-M, Parameters=355M2025.026.7
GPT2-SModel=GPT2-S, Parameters=127M2025.024
DiffuGPT-MModel=DiffuGPT-M, Parameters=355M2025.023.8
DiffuGPT-SModel=DiffuGPT-S, Parameters=127M2025.022
SEDD-MModel=SEDD-M, Parameters=424M2025.021.8
SEDD-SModel=SEDD-S, Parameters=170M2025.021.5
Plaid1BModel=Plaid1B, Parameters=1.3B2025.021.2