Using 138 low column density (N$_{\rm HI}=10^{12.5-16.5}$ cm$^{-2}$) Ly$\alpha$ “forest” absorbers discovered by HST at $z \leq 0.15$ and nearly one million low-$z$ galaxy redshifts from on-going surveys, we find that:

1. Although nearly 80% of Ly$\alpha$ forest absorbers are found within galaxy filaments, very few of them can be identified plausibly with bright ($\geq\,0.1\;{\rm L}^*$) galaxies. Either these absorbers are not related to any one individual galaxy or the individual galaxy to which they are related is fainter than $0.1\;{\rm L}^*$ (M$_B \geq -17$). O VI absorbers (H I+O VI) are found exclusively in galaxy filaments at median distances half as far away from the nearest galaxy as the Ly$\alpha$ absorber population from which they were selected (N$_{\rm HI}=10^{13.5-16.5}$ cm$^{-2}$).

2. Our work on individual absorber-galaxy associations at the lowest redshifts finds cases where low column density (N$_{\rm HI} = 10^{13.5-16.5}$ cm$^{-2}$) metal-enriched absorbers appear to be due to unbound winds from dwarf galaxies while higher column density (N$_{\rm HI}=10^{17.3-20}$ cm$^{-2}$) absorbers are found in the bound halos of massive galaxies. Therefore, both the statistics and a few well-studied examples lead us to propose that the low-$z$ IGM (and the Ly$\alpha$ forest absorbers in particular) are enriched in metals by outflowing winds from dwarf galaxies.