The paper "Class-Incremental Learning with CLIP: Adaptive Representation Adjustment and Parameter Fusion" addresses the challenge of class-incremental learning (CIL) using pre-trained vision-language models like CLIP. The authors propose a method called Adaptive Representation Adjustment and Parameter Fusion (RAPF) to reduce forgetting and improve model stability. RAPF uses textual features to adjust the representations of old classes affected by new classes, enhancing the separation of neighboring categories. Additionally, a decomposed parameter fusion strategy is employed to further mitigate forgetting during adapter module fine-tuning. The method is evaluated on several benchmarks, showing state-of-the-art results. Key contributions include the use of textual features to guide category separation and a parameter fusion approach that maintains stability without increasing the number of parameters. The paper also includes ablation studies and comparisons with existing methods, demonstrating the effectiveness of RAPF in reducing forgetting and improving performance.The paper "Class-Incremental Learning with CLIP: Adaptive Representation Adjustment and Parameter Fusion" addresses the challenge of class-incremental learning (CIL) using pre-trained vision-language models like CLIP. The authors propose a method called Adaptive Representation Adjustment and Parameter Fusion (RAPF) to reduce forgetting and improve model stability. RAPF uses textual features to adjust the representations of old classes affected by new classes, enhancing the separation of neighboring categories. Additionally, a decomposed parameter fusion strategy is employed to further mitigate forgetting during adapter module fine-tuning. The method is evaluated on several benchmarks, showing state-of-the-art results. Key contributions include the use of textual features to guide category separation and a parameter fusion approach that maintains stability without increasing the number of parameters. The paper also includes ablation studies and comparisons with existing methods, demonstrating the effectiveness of RAPF in reducing forgetting and improving performance.