Text-to-Image Personalization on Concepts dataset benchmark leaderboard · SOTA2 Research

Text-to-Image Personalization on Concepts dataset

0.717CLIP-I ScoreContinual Alternating

Evaluation Results

Method	Links
Continual Alternating2026.05		0.717	0.287	0.504	0.515	0.475	0.51	0.27	9.529
SeqLoRA (Bilevel)2026.05		0.715	0.28	0.494	0.439	0.439	0.515	0.271	8.898
SeqLoRA (Bilevel)2026.05		0.714	0.279	0.488	0.445	0.446	0.512	0.271	8.828
SeqLoRA (Bilevel)2026.05		0.711	0.279	0.487	0.435	0.434	0.508	0.271	8.772
SeqLoRA (Bilevel)2026.05		0.71	0.287	0.502	0.509	0.475	0.513	0.271	9.539
SeqLoRA (Bilevel)2026.05		0.707	0.278	0.478	0.437	0.44	0.501	0.271	9.136
Continual Alternating2026.05		0.706	0.278	0.485	0.429	0.424	0.504	0.271	9.032
SeqLoRA (Bilevel)2026.05		0.705	0.277	0.484	0.431	0.428	0.507	0.271	8.904
SeqLoRA (Bilevel)2026.05		0.704	0.282	0.484	0.452	0.439	0.494	0.269	9.476
SeqLoRA (Bilevel)2026.05		0.695	0.284	0.474	0.427	0.425	0.478	0.27	9.171
SeqLoRA (Bilevel)2026.05		0.693	0.286	0.468	0.418	0.418	0.477	0.27	9.043
Mix-of-Show2026.05		0.677	0.278	0.44	0.439	0.411	0.447	0.272	8.847
LoRACLR2026.05		0.674	0.278	0.435	0.431	0.403	0.445	0.272	9.089
Orthogonal Adaptation2026.05		0.664	0.274	0.416	0.405	0.379	0.429	0.272	8.886