[JoeyBob]

So gem-diols are stabalized by withdrawing groups, correct? Would this mean a nitrile group would stabilize it while a hydroxyl group would destabilize it?

[hollytara]

Yes - gem diols are favored by electron withdrawing groups adjacent to the carbonyl. 

Gem-diols are also called aldehyde or ketone hydrates - for most aldehydes, the gem diol (hydrate) form is dominant in water solution, but with removal of water the equilibrium shifts back to the aldehyde.  Ketones show less hydrtae formation without special characteristics.

The classic "stable, isolatable" gem diols are chloral hydrate (CCl3CH(OH)2) and ninhydrin. 

cyano acetaldehyde will be more stable bu probably not enough to be isolated.

a hydroxy group is a pi donor but sigma withdrawing - the effect is less predictable for 2-hydroxyacetaldehyde (glycolaldehyde).  This is an interesting molecule - exists independently mostly as a dimer, but in water is mostly the hydrate.

[Babcock_Hall]

Some years ago I saw a paper on dihydroxyacetone, which indicated that the hydrate form was about 80% and the ketone form was about 20%.  I do not have the reference handy.

[hollytara]

I would imagine the dihydroxyacetone dimerizes about as well as glycolaldehyde - they can form 6 membered ring cyclic structures that are hemiacetals/hemiketals.

[Babcock_Hall]

Good point.  IIRC there is also some work on dihydroxyacetone phosphate, probably from the laboratory of Jeremy Knowles, on this equilibrium.