Yes, that's true. But are there arrangements of dark matter that are not holding visible matter galaxies together? Does dark matter appear anywhere else?
Yes, that's how we'd infer the presence of an overdensity of dark matter in the absence of visible matter : a background being lensed by an invisible foreground.
There are searches for this which are possible because Einstein rings are well understood, and those appear in particular alignments such that it would be hard to miss the foreground source if it were at all visible. Other lensing structures are subject to tracing error; that is, you lose the ability to distinguish between effects from alignment and effects from the distribution of foreground mass and the distance to the background visible objects.
One major problem is that dark matter overdensities need not be in exact alignments, or produce readily detectable Einstein rings. Moreover, the overdensities must not be that "over" or ordinary matter would have been drawn in, and we could see that. (Which would be neat as it would offer evidence on how early galaxies formed in the first place, and narrow out of contention lots of modified gravity proposals.)
There is a debate over the large scale structure of dark matter, with lots of ideas provoked by the X-ray clusters Abell 222 & 223 among others. There may be a "web-like" structure to dark matter where the filaments can be very large compared to the scales of individual galaxies but still quite narrow compared to the less dense spaces between them. These structures appear pretty commonly in simulations grounded in the standard cosmology, although it's not wholly clear why (they may be an artifact of numerical methods, for example).
There are ground- and space-based observatories in the pipeline that will over the next few years narrow the possibilities for various kinds of large scale structure of extra-galactic/extra-galaxy-cluster dark matter.
