It's probably going to be difficult if not impossible to offer any concrete guidance without more information.
- What are the inputs to the function in MATLAB (both what do they represent/how are they used and potentially what actual values are you passing into the function)?
- How are you deploying your function to .NET? The documentation states there are (at least, I'm not that familiar with the .NET deployment functionality) two different options.
- How are you invoking the .NET assembly in your .NET code? Showing exactly what you're passing in (type and value) would be helpful.
- How different is "totally different"? Are we talking about an error near floating-point roundoff or are we talking many, many orders of magnitude greater than that?
I suspect that it would probably be easier to debug if you contact Technical Support directly using this link, show them the exact code you're running, and work with them to determine the cause of the difference in the results.
