An obscure fact I never knew: All of the lanthanides except promethium exhibit a divalent state.

[eppur_se_muova]

6. Yes, apparently it was electrides that led Dye et al. to try making alkalides originally.

IIRC, there was an article on this in Scientific American when I was in high school -- or maybe not. I just know I came across it at a time when I knew enough about chemistry to find such a thing almost unthinkable. Alkali metals were always cations! But someone thought otherwise. If I search my old notes I might find a record of a lecture from grad school on cryptands, so it's been brought to my attention a couple of times.

https://en.wikipedia.org/wiki/Alkalide

Then there's this:

Rare oxidation states

Less common oxidation states of gold include ?1, +2, and +5.

The ?1 oxidation state occurs in aurides, compounds containing the Au? anion. Caesium auride (CsAu), for example, crystallizes in the caesium chloride motif;[39] rubidium, potassium, and tetramethylammonium aurides are also known.[40] Gold has the highest electron affinity of any metal, at 222.8 kJ/mol, making Au? a stable species.[41]

https://en.wikipedia.org/wiki/Gold#Rare_oxidation_states

and this:

https://www.sciencedirect.com/science/article/pii/S129325580500230X

Apparently both BaPt and Cs2Pt are known.

Oh, and Pekka Pyykö (I love that name!) has predicted that Oganesson - element 118, a "noble gas" element, should form an anion - good thing IUPAC gave that name the nonmetallic -on ending!