GFrames: Gradient-Based Local Reference Frame for 3D Shape Matching. Vector of Locally Aggregated Embeddings for Text Representation. How well do NLI models capture verb veridicality?. QUARCH: A New Quasi-Affine Reconstruction Stratum From Vague Relative Camera Orientation Knowledge. Original Semantics-Oriented Attention and Deep Fusion Network for …