Keep in mind that UHF signal powers (especially) will be far below those predicted by TVFool when this close to the towers. The simulation assumes that all transmitting antennas radiate equally in the vertical plane which they do not. At only several miles from Sutro, the signal powers will be a mere fraction of those in the main lobe. For example, KPIX transmits about 686 kW in this OP's direction in its main lobe. However, the angle of depression is about 10°. KPIX's FCC application (
https://licensing.fcc.gov/cdbs/CDBS_...um=1&exhcnum=1) happens to include the calculated vertical pattern on page 9 of the engineering exhibit. According to that plot, voltage gain is about 0.1 the maximum at 10° below horizontal. Convert that to a power gain by squaring that and then multiplying by the ERP in the main lob and the result is ~6860 watts. This is a relatively extreme example as KPIX uses a very high gain antenna. VHF stations and FM are much lower gain with much more ERP at lower depression angles so their signals are going to be closer to the calculated (in TVF) signal powers.
Some attenuation is likely to be helpful as it will also reduce any undesired signals that might be present and causing issues.
Move the antenna several feet and try again. Repeat as needed.