The notion of "positions" is extremely important to the bond market (and to any market for that matter).  At the most basic level, a LONG position seeks to profit from prices rising (or yields falling) and a SHORT position seeks to profit from prices falling (or yields rising).  

Quite often, the structure of accumulated positions in the Treasury market acts as its own driver of momentum.  The key reason for this is the very logical practice of traders protecting themselves from undue losses with automatic trades unsurprisingly known as "stop losses" or simply "stops." 

For instance, if the market is overly long, and if a big chunk of longs were set-up around a specific trading level, if yields happen to move higher than that trading level (or something close to it), we're likely to see those longs get "stopped out"--that is, forced to sell due to their stop-loss levels.  This was certainly behind some of the nastier momentum seen on Tuesday as defensive longs had hoped 3.04% would prove to be a ceiling once again.  When it was broken, those longs had to cut bait.

The opposite of forced liquidation of longs (as described above) is a "short squeeze."  If you were able to understand the explanation above, then you already understand a short squeeze.  It occurs when a relatively big chunk of short trades are made near a certain level followed by briefly rising rates and then a rally back to those levels.  Like the longs, short positions also have stop-losses, but because a short position is technically a SALE of bonds today with an agreement to buy back at a future price, the short position is closed by BUYING bonds.

In short (pun intended), there have almost certainly been new shorts taken out as the most recent selling spree materialized.  Traders who were on the fence likely got pulled into the bearish momentum and put in new short bets at recent yield plateaus--especially where those plateaus coincide with volume.  The following chart shows two of them.  The implication is that if yields manage to break below these levels, there would likely be more buying to follow in the form of short-covering.  Short-covering pushes yields even lower, thus triggering the stop-loss levels for more and more shorts.  That's a short squeeze.  It's a domino effect to be sure.

The only catch is that a short squeeze is no basis for a sustained rally.  So, unless new "longs" are also entering the market, the rally could be sharp but temporary.  Unfortunately, the data published on trading positions always lags the market by a few days.  For instance, we'll get Tuesday's position report today, which doesn't do us much good considering we're most interested in what's happened since then.