Question on the pseudo code of arxiv paper

[treya-lin]

Hi thanks for your great work. I am reading the paper and noticed a bug which I assume to be a typo.

In the arxiv paper (https://arxiv.org/abs/2303.00747), Listing1 shows the pseudo code for smoothing. I noticed that the min-cut modification implementation seems incorrect. instead of

sc = scores[t]
...
pdx = i + max_len //2
qdx = i + max_len
min_sp = argmin(scores[pdx:qdx])
segs.append((start, pdx+min_sp))

shouldn't it be like this?

sc = scores[i]
...
pdx = start + max_len //2
qdx = start + max_len
min_sp = argmin(scores[pdx:qdx])
segs.append((start, pdx+min_sp))

I also checked the code in this repo, which seems to be consistent with the corrected version. Correct me if I misunderstood it?