PEDRO: Parameter-Efficient Fine-tuning with Prompt DEpenDent Representation MOdification | Read Paper on Bytez