Adversarial Attacks on AI-Generated Text Detection Models: A Token Probability-Based Approach Using Embeddings

Open in new window