Understand prstat result

Here is top 10 of prstat result, what is that mean by load averages, is this consider high or very low ?


testDummy> prstat -s cpu -n 10
   PID USERNAME  SIZE   RSS STATE  PRI NICE      TIME  CPU PROCESS/NLWP
 27365 root     7032K 1632K cpu1     0    0 8080:19:5  12% util_sybase.pl/1
  3411 root      181M  180M cpu6     0    0   1:05:17  12% logadm/1
 16109 sybase     10G   10G cpu2    12    0 2471:33:4 5.2% dataserver/8
 16097 sybase     10G   10G sleep   59    0 2364:55:0 4.8% dataserver/17
  2415 sybase     11M 7720K sleep   50    0   0:19:16 3.1% sybmultbuf/1
 16108 sybase     10G   10G sleep   59    0 2344:51:3 1.1% dataserver/8
 14082 sybase    511M  493M sleep   59    0 1270:24:4 0.8% dataserver/17
  2263 sybase    535M  516M sleep   59    0  53:33:27 0.3% dataserver/17
 19614 root       16M   12M sleep   56    1  11:23:46 0.1% perl/1
  2416 sybase   3032K 2584K sleep   59    0   0:00:19 0.0% sybmultbuf/2
Total: 137 processes, 1303 lwps, load averages: 3.50, 3.81, 3.65