The GTSAM 3.1.0 release has the following new features and improvements:

• Smart Projection Factors, which exploit the underlying structure of the optimization problem to enhance efficiency and robustness to degenerate configurations. Presented at ICRA: Eliminating Conditionally Independent Sets in Factor Graphs: A Unifying Perspective based on Smart Factors. Please see SFMExample_SmartFactor.cpp for a usage example.
• Linear Approximation for Graph Optimization (LAGO), published in IJRR: A Fast and Accurate Approximation for Planar Pose Graph Optimization
• Matlab toolbox: Substantial speedup when plotting 2D poses. 
• Various small bugfixes