In reality, some things are more important to you than others. To model this fact for your agents, you can set Significance values for each perceived object using the Steering Tag. For a steering behaviour, UseSignificance determines if this feature is activated or if the default value of 1 is used.

An example case: We have got a scene where an agent faces different types of enemies. For example, a puny thief and a mighty knight in a shiny armor which looks like it has awesome stats. The agent utilizes a Seek behaviour which writes values to a danger objective, thus, avoiding directions to these enemies. Now, if you want your agents to be suicidal (because you can), you would assign a high Significance to the knight characters and a small Significance to the thief characters. Most likely, you would do the exact opposite so that your agent might survive by avoiding the knight characters. A similar example setup can be seen in the following figure.

Figure 1: The agent has a Seek behaviour which perceives all four green circles. All circles have the same distance to the agent (indicated by the blue circle). Thus, without different Significance values, the resulting magnitudes which are visualized by the indicator gizmo would be equal for all objects. However, in this example, the Significance of the objects corresponds to the values which are displayed inside of the circles.