List your company

Massive Multitask Language Understanding on MMLU (Accuracy, Delta, and Forgetting Profile)

53AccuracyBase
20.44828.89937.3545.801Apr 4, 2026

Evaluation results

MethodLinks
BaseBackbone=Qwen2.5-72B2026.0453——
Standard LoRABackbone=Qwen2.5-72B, Campaign=1, Number of seeds=32026.04530—
AletheiaBackbone=Qwen2.5-72B, Campaign=1, Number of seeds=32026.045300
Standard LoRABackbone=Qwen2.5-14B, Campaign=1, Number of seeds=32026.0452.20.8—
AletheiaBackbone=Qwen2.5-14B, Campaign=1, Number of seeds=32026.04520.70.2
BaseBackbone=Qwen2.5-14B2026.0451.3——
BaseBackbone=Llama-3.1-8B2026.0450.8——
Standard LoRABackbone=Qwen2.5-7B, Campaign=1, Number of seeds=32026.0450.31.3—
BaseBackbone=Qwen2.5-7B2026.0449——
AletheiaBackbone=Llama-3.1-8B, Campaign=1, Number of seeds=32026.04491.80.2
Standard LoRABackbone=Llama-3.1-8B, Campaign=1, Number of seeds=32026.0448.82—
AletheiaBackbone=Qwen2.5-7B, Campaign=1, Number of seeds=32026.0448.20.82.2
BaseBackbone=Qwen2.5-1.5B2026.0446——
Standard LoRABackbone=Qwen2.5-1.5B, Campaign=1, Number of seeds=32026.0445.20.8—
AletheiaBackbone=Qwen2.5-1.5B, Campaign=1, Number of seeds=32026.0444.21.81
AletheiaBackbone=Mistral-7B, Campaign=1, Number of seeds=32026.0441.311.5
BaseBackbone=Mistral-7B2026.0440.3——
Standard LoRABackbone=Mistral-7B, Campaign=1, Number of seeds=32026.0439.80.5—
BaseBackbone=Qwen2.5-0.5B2026.0437.3——
AletheiaBackbone=Qwen2.5-0.5B, Campaign=1, Number of seeds=32026.0436.70.70.5
Standard LoRABackbone=Qwen2.5-0.5B, Campaign=1, Number of seeds=32026.0436.21.2—
AletheiaBackbone=Phi-3.5-mini, Campaign=1, Number of seeds=32026.0426.23.20.7
Standard LoRABackbone=Phi-3.5-mini, Campaign=1, Number of seeds=32026.0425.52.5—
BaseBackbone=TinyLlama-1.1B2026.0423.5——
AletheiaBackbone=TinyLlama-1.1B, Campaign=1, Number of seeds=32026.0423.501.8
BaseBackbone=Phi-3.5-mini2026.0423——
BaseBackbone=Llama-3.1-70B2026.0422.7——
Standard LoRABackbone=Llama-3.1-70B, Campaign=1, Number of seeds=32026.0422.70—
AletheiaBackbone=Llama-3.1-70B, Campaign=1, Number of seeds=32026.0422.700
Standard LoRABackbone=TinyLlama-1.1B, Campaign=1, Number of seeds=32026.0421.71.8—