3 results about "Dead code" patented technology

The present invention relates to a method performed by a system (110) operable to configure a chemical separation system (120), the method comprising: receiving system hardware configuration data, wherein the system hardware configuration data is at least indicative of fluid manipulation modules (121_1-12N_M) of the chemical separation system (120), desired functionality of the fluid manipulation modules, and a fluid network (210) comprising fluid couplers configured to fluidly couple the fluid manipulation modules (121_1-12N_M), identifying a first set of function blocks to provide the indicated fluid manipulation modules with the desired functionality, generating a second set of function blocks by purging a third set of function blocks from the first set of function blocks to eliminate dead code using the configuration data, sending the second set of function blocks to the system (120).

The embodiment of the application provides a code processing method, device, computer equipment, computer readable storage medium and computer program product, and belongs to the technical field of software development. The code processing method comprises the following steps: before performing a target change operation on a target code library, scanning the target code library by using a dead code detection tool to obtain a baseline dead code set; in the case that the target change operation on the target code library is completed, scanning the target code library by using the dead code detection tool to obtain a current dead code set, the storage structure of the current dead code set being consistent with the storage structure of the baseline dead code set; determining the difference between the current dead code set and the baseline dead code set, and determining the incremental dead code of the operation based on the difference; and performing invalid code cleaning based on the incremental dead code. The technical scheme of the embodiment of the application can effectively identify the newly added dead code, and improve the efficiency of dead code cleaning.

This invention discloses a robustness enhancement method for large-scale code models based on context processing and post-training, selecting the enhancement method according to the type of model to be enhanced. For post-training enhancement, adversarial perturbation samples are first generated from the given source code using equivalence semantic transformation, greedy search, and simulated annealing. These perturbation samples are then input into the post-training framework to fine-tune the target large-scale language model. For context enhancement, a backup dataset is generated using backup code. The optimal subset of examples most semantically similar to the current problem is selected by calculating similarity. Dead code is then normalized, and the optimal subset of examples is adaptively selected. Contextual prompts are dynamically adjusted based on task complexity and input into the target large-scale language model. Finally, evaluation metrics and the robustly enhanced target large-scale language model are output. This invention supports multiple mainstream programming languages ​​and both open-source and closed-source models, exhibiting good versatility and scalability.