PDN 中的 PATH 分数计算与论文中的公式不符

https://github.com/alibaba/EasyRec/blob/d965fb796ca834285b82dec071128aec3a82c584/easy_rec/python/model/pdn.py#L43-L45

代码中实现的是：

```
trigger_out = tf.exp()
sim_out = tf.exp()
logits = tf.multiplay(sim_out, trigger_out)
```

相当于是：

$$ PATH_{uji} = e^{t_{uj}} e^{s_{ji}} $$

论文中的合并公式（公式(10)）是：

$$
PATH_{uji} = MEG(t_{uj}, s_{ji}) = \ln(1 + e^{t_{uj}} e^{s_{ji}})
$$

Start by reading easy_rec/python/model/pdn.py lines 43-45 and compare the calculation with formula (10) in the referenced paper. Look for existing PDN model tests or validation around this calculation; done means the PATH score follows the paper’s formula and the relevant behavior is validated.

Written by the indexing model from the issue text.