Checking device array in CUDA

So I have a Particle System that I'm trying to convert over into CUDA, but am having some trouble actually checking if it's doing what I want it to do.

I am trying to allocate a temporary array in my constructor, initialize it and then pass it into a device array for the kernel calls. I'm not very familiar with CUDA and don't know of a better way to check if my kernels are doing what I want, so for now I was trying to copy the device array back into a host array and just checking the values after a single time step. Unfortunately, all I'm getting is garbage data right now.

Thanks.

In my header file:

Particle tempParticles[numParticles];
Particle* particles;

Constructor:

ParticleSystem::ParticleSystem() {
    cudaMalloc((void**)&particles, numParticles * sizeof(Particle));
    int count = 0;
    //for loops
        tempParticles[count].newPos = glm::vec3(i, j, k);
        count++;
    //end for

    cudaMemcpy(particles, tempParticles, numParticles * sizeof(Particle), cudaMemcpyHostToDevice);
}

Update Wrapper for cuda:

void ParticleSystem::updateWrapper() {
    update(particles);
    cudaMemcpy(tempParticles, particles, numParticles * sizeof(Particle), cudaMemcpyDeviceToHost);
}

Example of my cuda update method:

__global__ void predictPositions(Particle* particles) {
    int index = threadIdx.x + (blockIdx.x * blockDim.x);
    if (index > NUM_PARTICLES_C) return;

    //predict position x* = xi + dt * vi
    particles[index].newPos += particles[index].velocity * deltaT;
}

void update(Particle* particles) {
        predictPositions<<<dims, blockSize>>>(particles);
}