What is causing this code?

You may have a manifold vacuum leak you can test for with propane or a smoker but you may also have a sticking injector.
Those are the common ones. MAF sensors etc...many things could go wrong but are less likely.
ECU compensation for lean frequently convolutes diagnosis.
Dump in some GOOD injector cleaner (not $2/bottle) and run a tank. You may get lucky and see fuel trims come down.
This fixed a similar problem for me in another engine. (and I have the unused rebuild parts to prove it )
BTW: trims change in milliseconds/seconds...not day/weeks. The ECU monitors engine/exhaust conditions and adjusts with trims rapidly.
The freeze frame trims you showed were recorded at the time the code was thrown.
You can monitor trims in real time if you have a simple OBDII data logger (app and phone or laptop works well) but I'd try the simple fix first...strong injector cleaner.
Simple things first.

Edit: Looking closer at what you posted (first showed just B1..2nd showed B2..I see you were dealing with a small screen), B1 and B2 seem to have a problem so one injector may not be the issue. Cleaner can't hurt and has a chance, but now i think less likely.

 

The idea is right but that's not even close on time and how they are determined.

STFT is fuel correction that happens many times a second. This is usually split into 4+ load ranges like A,B,C,D or similar for various amounts of manifold pressure/MAF value ranges. STFT can react quickly to changes.

LTFT is typically an avg of the STFT, maybe with other inputs or calcs. How these are applied to fuel schedules, tables, etc beyond the blanket application to the relevant load cells is a little fuzzy to me and I think that might be because it varies by ECU logic like between manufacturers, EFI systems, etc

On various vehicles that I have seen data from, fuel trims under 5% is common for normally operating engines. My completely stock 2GR-FE stays around +3%. STFT can swing more, but they should settle within a few % of 0 when conditions stabilize.

 

You're absolutely correct but my intent was to show a Relative time frame to keep the explanation simple. I should have clarified that "STFT reacts quickly to changes." It is data in real time.

Thanks for providing the "Long Version!"

 

Thanks to both of you!

 

Well sprayed the air flow sensor well and checked all air box and hoses with no positive results. Going to try the propane method today.

 

I remember getting a code similar to that, left bank lean, found out it was a hose that came off from underneath the air box. might check it out.