Difference in elevation in Sentinel SLC DEM and ALOS/SRTM3o

I have a Sentinel SLC data generated DEM ! If I compare it with SRTM 30 m or ALOS PAR DEM. There is a difference in maximum elevation around 2000 m.

Any sugestions?

I think the extreme values do not matter the most, because they are likely to be caused by InSAR errors, such as decorrelation, phase unwrapping errors, atmosphere ect…
Have a look here for an overview of possible error sources: https://eo-college.org/resources/insar_errors/

It is important that the general pattern of your DEM makes sense. And you need high coherence for the DEM to be of good quality which can vary within the image quite much.

Can you post screenshots of the DEM and the coherence image (including scale)?
Can you post the names of the image products you used for the DEM generation?

New Microsoft PowerPoint Presentation.pptx (1.5 MB)

S1A_IW_SLC__1SSV_20141026T124702_20141026T124729_003001_0036B0_6419 - Master Image
S1A_IW_SLC__1SSV_20141107T124714_20141107T124744_003176_003A7E_AD9F - Slave Image

thank you. I see that coherence is very low over wide parts of the study area, most likely because of vegetation cover. In these regions, no usable information on elevation can be retrieved.

Some thoughts on this:
Did you try another image pair? Sometimes atmospheric conditions are strongly disturbing the phase information.
Was your coregistration successful?

I tested your scenes and retrieve acceptable interferograms
before unwrapping:

unwrapped:

I am trying other pair also! Is anything wrong with the way I am exporting it?

I cannot judge based on these two images, sorry.