Alibaba's PoS: Beyond memory for long-horizon agents
An inference-time framework that maintains explicit belief states, detects when agents stall (Belief Trapping), and recovers with targeted constraints. Achieves SOTA across 4 benchmarks, with gains up to 37.89%.
We have computational social science postdoc positions here at Pitt. Appreciate if you can forward it to your group and anyone who may be interested in the position. Thanks! https://t.co/vOJoJY8YmB
Excited to share that our work "Iterative Discriminant Tensor Factorization for Behavior Comparison in Massive Open Online Courses" has been accepted as a full paper at @TheWebConf#www2019, with @rhodiuslin@peterpaws, Xi Liu, and @j_barria!