I will survey recent results on scaling limits of self-repelling random walks and diffusions which are pushed by the negative gradient of their own occupation time measure, towards domains less visited in the past. The typical examples are the so called 'true (or myopic) self-avoiding walk' or the 'self repelling Brownian polymer process'. It is proved that in three and more dimensions the processes scale diffusively, in two dimensions (this is the the critical dimension of the phenomenon) multiplicative logarithmic corrections are valid.