15:22 <vaibhavp[m]> Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method you have passed target/output to the second parameter, which should be the delta of the next layer(i.e. backward of the next layer)? Is it some kind of mathematical trick because I have seen this in many places... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/3e60cb39b251b2745ca3adffe3358212cebb61e0>)

15:23 <vaibhavp[m]> * Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method you have passed target/output(see code/link below) to the second parameter, which should be the delta of the next layer(i.e. backward of the next layer)? Is it some kind of mathematical trick because I have seen this in many places but it makes no sense to me? ([Line 147 in

15:23 <vaibhavp[m]> ```

15:23 <vaibhavp[m]> ann\_test\_tools.hpp](https://github.com/mlpack/mlpack/blob/master/src/mlpack/tests/ann/ann_test_tools.hpp#L147))

15:23 <vaibhavp[m]> ...(truncated)

15:23 <vaibhavp[m]> * Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method you have passed target/output (see code/link below) to the second parameter, which should be the delta of the next layer(i.e. backward of the next layer)? Is it some kind of mathematical trick because I have... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/5ae0ca5bede151c355e2cbb1a2570690e1573c5d>)

15:24 <vaibhavp[m]> * Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method you have passed target/output (see code/link below) to the second parameter, which should be the delta of the next layer(i.e. backward of the next layer)? Is it some kind of mathematical trick because I have... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/1a14942100345d633b8642513d206b79a478b356>)

15:24 <vaibhavp[m]> * Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method(Link 147) you have passed target/output (see code/link below) to the second parameter, which should be the delta of the next layer(i.e. backward of the next layer)? Is it some kind of mathematical trick... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/91a815294933b9b844e43884773d1c40cc4b3afe>)

15:24 <vaibhavp[m]> * Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method(Line 147) you have passed target/output (see code/link below) to the second parameter, which should be the delta of the next layer(i.e. backward of the next layer)? Is it some kind of mathematical trick... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/fa4c652dc8a9d0ed337cf5efaf630380d934a6d7>)

15:35 <vaibhavp[m]> * Also zoq , Can you explain how does the _JacobianPerformanceTest_ work, because to the _Backward_ method(Line 147) you have passed target/output (see code/link below) to the second parameter, which should be the delta of the next layer(i.e. backpropagated error)? Is it some kind of mathematical trick because I... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/ef97e740d99bab3436574861ec5bbad3b276b015>)