1. The scale-dependent part of additional bias (A(k), in Eq.(4) in

1208.1491) is plotted in the attached figure "akfig.pdf". It seems that additional bias contribute more power on large scales than small scales. How do we understand this?

akfig.pdf

2. If A(k) I plotted is correct, then b

_{tot}(k)=b+ Delta b(k) is shown in "btotfig1.pdf". Just want to make clearer about the relations between different choices of parameter values and the shape of the b

_{tot}. There are four cases:

btotfig1.pdf

(a) bL>0,(b>1) and fNL>0: purple line. This means that galaxy is more correlated with matter fluctuations, and small scale galaxy fluctuation is correlated with large scale fluctuations, so b

_{tot}at large scale becomes large. Is this right?

(b) bL<0 (b<1) and fNL>0: green line. Galaxy initially is anti-correlated with matter fluctuations (bL<0), therefore after evolution (extra 1 factor), Galaxy is less correlated with matter fluctuation and matter auto-correlation (0<b<1). fNL>0 means small scale gravity potential is correlated with large scale potential fluctuation, right? But how to understand these two things combined to give the large scale suppression?

In "pgfig1.pdf", we plot P

_{g}(k), you can find that the green line and

purple line are combined with each other on k<10

^{(−3)}. Why these two sets of parameter choice give same power on very large scale?

pgfig1.pdf

(c)bL>0(b>1) and fNL<0: brown line. Why this set of parameter choice give a suppression on large scales.

(d) bL<0(b<1) and fNL<0: blue line. galaxy is initially anti-correlated

with matter fluctuation (bL<0), and small scale and large scale potential fluctuations are anti-correlated as well. Why this set of parameter choice results in a boost on large scales as well?

Thank you guys!