It seems you are making several beginner's mistakes in your understanding and application.
You are using the LM2956 5.0 which is the 5 V version. This will output a regulated 5.0 V subject to meeting certain conditions including the required input voltage and current and having a load drawing between the rated minimum and maximum.
The 3 A maximum output rating of the supply means that it can guarantee 5 V out while delivering 3 A into a load. This means that the minimum load resistance is \$ R = \frac {V}{I} = \frac {5}{3} = 1.7 \ \Omega; \$. You do not test power supply output current by short-circuiting its output with an ammeter. To do so risks destroying the meter or, at best, blowing the fuses. In addition the reading is of very little use as that current is delivered when there is no voltage at the output (due to the short-circuit) so you can't power anything with it.
Figure 1. LED current versus voltage curves for various colours. Source: LED IV curves.
Next. You are showing a dummy load of an LED connected directly across the output. LEDs have a forward voltage, Vf, usually given when a current of 20 mA is run through the LED. You can see from the graphs that if you connect an LED directly to a 5 V supply that for most of the colours the current drawn is off the scale and, generally, the LED will glow very brightly for a very short time. If your simulator doesn't show that your LED is destroyed then it has a poor LED model. See LED resistor calculation for more on how to limit the current with a resistor.
How to test your circuit:
simulate this circuit – Schematic created using CircuitLab
Figure 2. Correct PSU monitoring technique.
The ammeter is always wired in series with the load. Voltage measurements are taken across the load (or "in parallel" with the load).
Repeat your simulations while increasing the load resistance from 1.7 Ω and see when the regulator starts to misbehave.
Tip: turn off the grid before taking screengrabs. It makes your diagrams easier to read.