About Figure 3(d)

[xizaoqu]

Hi~I found that in cases like Figure 3(a), one policy only applies to one maze, and needs references for reconstruction. Do cases in Figure 3(d) need references, or can be generalized to multiple scenes? By the way, have scripts for Figure 3(d) been released in this repo?

[jannerm]

What do you mean by references for reconstruction?

[xizaoqu]

What do you mean by references for reconstruction?

Sorry for not making it clear. I mean to generate trajectories for a specific maze, we first need some ground truth trajectories (reference) (Figure 3(a)). It seems plausible to generate trajectories with a reward map (Figure 3(d)) and without reference. Did Figure 3(d) do in this way?

[jannerm]

The training dataset for those figures consists of random-walk trajectories in an enclosed square.