Recently, I found the firmware of Tenda is available to the public, thus I started to seek vulnerabilities in it, mostly focused on buffer overflow & rce. After hours and days of searching, examining, and reproducing; I was finally able to find 7 0day (mostly buffer overflows) from the Tenda

BERT * BERT is pretrained with two objectives: masked language modeling and next-sentence prediction * [SEP] [CLS] Intro & Coda * [CLS] classification input label The input embeddings are passed through multiple encoder layers to output some final hidden states. * To use the pretrained model for text classification, add a sequence classification head