Dialectal alignment estimation on ELI5 informal benchmark leaderboard · SOTA2 Research

Dialectal alignment estimation on ELI5 informal

0.77Percentage of AmE GenerationsGPT-4o

Evaluation Results

Method	Links
GPT-4o2026.04		0.77	0.82
Gemini-2.0-flash2026.04		0.7533	0.82
Claude-3.7-sonnet2026.04		0.7333	0.86
DeepSeek-V32026.04		0.7333	0.84
StableLM-2-1.6B2026.04		0.7333	0.77
Mistral-Small-24B2026.04		0.7233	0.82
Velvet-2B2026.04		0.71	0.8
Llama-3.3-70B2026.04		0.69	0.79
Gemma-3-27B2026.04		0.6833	0.83
Velvet-2B2026.04		0.6767	0.8
StableLM-2-1.6B2026.04		0.67	0.77
Falcon3-7B2026.04		0.66	0.81
Falcon3-7B2026.04		0.6333	0.77
Gemini-2.0-flash2026.04		0.4233	0.8
DeepSeek-V32026.04		0.4047	0.76
Mistral-Small-24B2026.04		0.3867	0.76
Gemma-3-27B2026.04		0.38	0.78
Claude-3.7-sonnet2026.04		0.3767	0.78
GPT-4o2026.04		0.3467	0.78
Llama-3.3-70B2026.04		0.3	0.76