Minor truncation in projected maps

Hi There,
I am generating gv.Points maps and keep running into truncation issues at the edges with projected data. It is usually very minor, but in some cases cuts enough to be noticeable, and tends to be magnified a little using panel.

I have not been able to find a way to pad or change limits a little to avoid this.
I have tried using various .opts() including xlim, ylim, padding, frame_width, frame_height, aspect and also messed with plot hooks in panel to modify ranges.

Using ZoomOutTool with bokeh/panel achieves the effect, but only with a manual zoom out, which I could not make automatic.

A basic example with the Molleweide projection (usually cuts a little of edge):

import geoviews as gv
import geoviews.feature as ft 
from geoviews import opts
from cartopy import crs
from bokeh.plotting import show


# Problem is minor truncation, but it seems like there should be
# some way to size/pad/limit to avoid entirely
cities = [(-74.0, 40.7, 'NYC'), (116.4, 39.9, 'Beijing'), (-0.13, 51.5, 'London')]

cities_latlon = gv.Points(cities, vdims='City')
molleweide = crs.Mollweide()

# Projected does not seem to use the lim/padding args
    xlim=(-10, 370),
    ylim=(-100, 100),
plotit = cities_latlon * ft.coastline
render_plot = gv.render(plotit)

# Example in panel - same issue, just a little more so
import panel as pn

cmap_sel = pn.widgets.Select(
    options=options, name='Color Map', value='jet', description='Choose a Color Map',
bokeh_server = pn.Row(cmap_sel, plotit).show(port=12345)

So if there are options to avoid or reduce the change of these little truncations whether that be in geoviews, panel, or the projection / data itself, I am interested to know about them.
I am using:
holviews 1.18.3, geoviews 1.11.1, bokeh 3.3.4, and panel 1.3.8 on ubuntu linux with Brave as the browser.

Can you share a screenshot and point out where the truncations are?

Otherwise, padding=0.1 potentially could help.

I will post a screenshot showing the issue later today.
The truncation only happens on a gv.Points plot for me right now, when I generate an gv.Image from the same data, the padding=0.1 argument applied successfully, but seemingly not in the point plot, which I was using for some of the hover functionality.

I found a way to make it work, though am somewhat puzzled as to why.

Here is what I was getting, truncation on either side:

With this code:

land_point_temps = selected_gv.to(gv.Points, ['lon', 'lat'], ['air'])
land_point_temps = land_point_temps.opts(
    title="Monthly Average Land Surface Temperatures",
    tools=[zoom_in, zoom_out],
# Then plot it

But by adding a feature layer with otherwise the exact same code (any seemed to work):

land_point_temps * ft.coastline

It was sorted out:

Thanks for the screenshots; is it possible to zoom out in the first, or is it still cropped?

Zooming out in the first case shows the whole plot without the cropping effect. The cropping does not seem to be innate in that regard.