Bug in inference mode?

I found that, in the code, if hidden_states.shape[1] != 1: is used to detect prefill stage.

I thought this criteria sometimes goes wrong. In my experiments, the hidden_states.shape was:

• hidden_states.shape[1] = n: prefill
• hidden_states.shape[1] =1: when generating the 2nd token
• hidden_states.shape[1] =2: when generating the 3rd token
• hidden_states.shape[1] =3: when generating the 4th token
  ...

So I have two questions:

1. Can hidden_states.shape[1] consistently equals to 1 (instead of 1,2,3,...) after prefilling if I use some generating setting?
2. Would position_ids[0,0] == 0 be a robuster criteria than hidden_states.shape[1] != 1 to detect prefill?

Start by locating the inference code containing if hidden_states.shape[1] != 1: and trace how hidden_states and position_ids are produced during generation. Compare the observed shapes under relevant generation settings, then document which signal reliably identifies prefill and define the resulting code change or recommendation.