In this paper, a new chaotic system is proposed based on mixing three‐dimensional Chen chaotic system with a chaotic tactics. This new system is proved to be chaotic under Wiggins’ chaos definition and can generate chaotic sequences with high complexity. Furthermore, we propose a new pseudorandom bit generator (PRBG) based on this new system. A coding algorithm is used to make the sequences uniform. Both statistical and security tests are provided to show the generated sequences are with good randomness and high complexity to withstand attacks.

Given that the sequences generated by logistic map are unsecure with a number of weaknesses, including its relatively small key space, uneven distribution, and vulnerability to attack by phase space reconstruction, this paper proposes a new two‐dimensional mutual coupled logistic map, which can overcome these weaknesses. Our two‐dimensional chaotic map model is simpler than the recently proposed three‐dimensional coupled logistic map, whereas the sequence generated by our system is more complex. Furthermore, a new kind of pseudorandom number generator (PRNG) based on the mutual coupled logistic maps is proposed for application. Both statistical tests and security analysis show that our proposed PRNG has good randomness and that it can resist all kinds of attacks. The algorithm speed analysis indicates that PRNG is valuable to practical applications.

At present, deep neural networks are at risk from backdoor attacks, but natural language processing (NLP) lacks sufficient research on backdoor attacks. To improve the invisibility of backdoor attacks, some innovative textual backdoor attack methods utilize modern language models to generate poisoned text with backdoor triggers, which are called feature space backdoor attacks. However, this article find that texts generated by the same language model without backdoor triggers also have a high probability of activating the backdoors they injected. Therefore, this article proposes a multistyle transfer-based backdoor attack that uses multiple text styles as the backdoor trigger. Furthermore, inspired by the ability of modern language models to distinguish between texts generated by different language models, this article proposes a paraphrase-based backdoor attack, which leverages the shared characteristics of sentences generated by the same paraphrase model as the backdoor trigger. Experiments have been conducted to demonstrate that both backdoor attack methods can be effective against NLP models. More importantly, compared with other feature space backdoor attacks, the poisoned samples generated by paraphrase-based backdoor attacks have improved semantic similarity.