List your company

Reading Comprehension on Belebele EN

75.33AccuracyLayerMoE
23.714837.114950.51563.9151Dec 3, 2025Dec 18, 2025Jan 2, 2026Jan 18, 2026Feb 2, 2026Feb 17, 2026Mar 5, 2026

Evaluation results

MethodLinks
LayerMoETarget Language=Turkish, Backbone=Llama-3.2-3B, Number of Experts=842026.0375.33
LayerMoETarget Language=Hungarian, Backbone=Llama-3.2-3B, Number of Experts=842026.0375.33
NeuronMoE-S1Backbone=Llama-3.2-3B, Stage=Stage 1, #Experts=49 (-42%)2026.0375.33
NeuronMoEBackbone=Llama-3.2-3B, Stage=Stage 2, #Experts=49 (-42%)2026.0375.11
NeuronMoETarget Language=Hungarian, Backbone=Llama-3.2-3B, Number of Experts=47 (-44%)2026.0374.78
LayerMoEBackbone=Llama-3.2-3B, Stage=Stage 2, #Experts=842026.0374.78
NeuronMoETarget Language=Turkish, Backbone=Llama-3.2-3B, Number of Experts=50 (-40%)2026.0374.44
Llama-3.2-3BTraining Status=No training2025.1274.2
DenseTarget Language=Turkish, Backbone=Llama-3.2-3B2026.0374.11
DenseTarget Language=Hungarian, Backbone=Llama-3.2-3B2026.0374.11
Llama-3.2-3B DenseBackbone=Llama-3.2-3B, Stage=Dense, #Experts=-2026.0374.11
LayerMoE-S1Backbone=Llama-3.2-3B, Stage=Stage 1, #Experts=842026.0373.89
NeuronMoE-ENBackbone=Llama-3.2-3B, Stage=Stage 1, #Experts=37 (-56%)2026.0373.56
Llama-3.2-3BTraining Status=Continued pretraining, Tokenizer Strategy=prune+ext (16k, naive)2025.1266
Llama-3.2-3BTraining Status=Continued pretraining, Tokenizer Strategy=prune+ext (16k, ours)2025.1265
Llama-3.2-3BTraining Status=Continued pretraining, Tokenizer Strategy=default2025.1264.7
NeuronMoE-S1Backbone=Qwen-1.5-1.8B, Stage=Stage 1, #Experts=36 (-50%)2026.0358.67
LayerMoE-S1Backbone=Qwen-1.5-1.8B, Stage=Stage 1, #Experts=722026.0354.89
Llama-3.2-1BTraining Status=No training2025.1235.7
Llama-3.2-1BTraining Status=Continued pretraining, Tokenizer Strategy=prune+ext (16k, ours)2025.1227.1
Llama-3.2-1BTraining Status=Continued pretraining, Tokenizer Strategy=default2025.1226.9
Llama-3.2-1BTraining Status=Continued pretraining, Tokenizer Strategy=prune+ext (16k, naive)2025.1225.7