Abstract: Systems and methods provide for a prompt template to include private random strings that are used by a language model to reference specific tokens on which the language model has been trained. A number of random strings may be generated and inserted into a prompt template and assigned special token identifiers (IDs) in a tokenizer. The text data from the prompt template are tokenized to convert the random strings to the assigned special token IDs which are sent to the language model for inferencing. Based on the provided special token IDs, the language model may reference special tokens learned from training and then generates an inference. The random strings remain hidden during their lifetime and may be updated on-demand to ensure security.

Abstract: In one embodiment of the present invention, a GPU contains an authentication module at the front end, and a memory security engine and graphic memory interface at the backend. In one embodiment of the present invention, the memory security engine provides a privilege table. The programmable privilege table maps memory address ranges, and user IDs to privileges for accessing the memory address ranges. In one embodiment of the present invention, the memory security engine receives a memory access command along with an associated authenticated user ID. In one embodiment of the present invention, the memory security engine checks the authenticated user ID and address range against the privilege table. In one embodiment of the present invention, if the table indicates that the user has authorization for the particular read or write transaction to the graphic memory, the instruction is executed by the graphic memory interface.