I am using OpenGL to do some GPGPU computations through the combination of one vertex shader and one fragment shader. I need to do computations on a image at different scale
The GLSL 1.20 specification (section 8.7) states that fragment shaders cannot choose their own mipmap level (and that the texture*Lod function are only available in vertex shaders). If anything, you may be able to use the bias parameter to the non-Lod variants to change the mipmap level, but it can only change it relative to what that the card has already calculated for you.
I don't know if latter versions of GLSL may have changed that.