Diode attenuator circuit gives me an offset?

We built the circuit shown here and is one of the labs posted on the wiki site.

https://www.analog.com/en/analog-dialogue/studentzone/studentzone-august-2019.html

The circuit does attenuate the input signl but there is an unexplained dc offset in the output signal relative to the input signal.  Below is an image of the scopy results.

the offset is about 86mV below the input level.  The output is channel 1+ and the input is channel 2+.

any idea why?