I've done some testing on this already, & I've learned that the TMOUT variable in bash (& ksh) is the one that controls the session timeouts, but only for interactive login shells (e.g., when I do "ssh hostname"). Scripted sessions are not affected (e.g., when I do 'ssh hostname "./test.sh"' & the script simply does a "sleep 960" then at the end runs the date command & outputs it to a test file on the remote box).
My initial research told me to set "ClientAliveInterval" & "ClientAliveCountMax" to 300 & 3, respectively, in /etc/ssh/sshd_config, then restart the sshd service (we run Solaris 10 & 11). But I tested it & those variables didn't do the job (my interactive login session stayed up forever). So then I tried the TMOUT variable in bash, set it to 900 in ".bash_profile", & it worked (i.e., I got disconnected after 15 minutes of inactivity).
Having already answered my own question, I ask the community if someone knows of another way to configure SSH session timeouts aside from the TMOUT variable in bash & ksh (we only use bash in our environment). The variables I mentioned in sshd_config don't seem to have any effect whatsoever, based on my tests.