Link prediction is an important way to complete knowledge graphs (KGs), while embedding-based methods, effective for link prediction in KGs, perform poorly on relations that only have a few associative triples. In this work, we propose a Meta Relational Learning (MetaR) framework to do the common but challenging few-shot link prediction in KGs, namely predicting new triples about a relation by only observing a few associative triples. We solve few-shot link prediction by focusing on transferring relation-specific meta information to make model learn the most important knowledge and learn faster, corresponding to relation meta and gradient meta respectively in MetaR. Empirically, our model achieves stateof-the-art results on few-shot link prediction KG benchmarks.

Crystalline optical cavities are the foundation of today's state-of-the-art ultrastable lasers. Building on our previous silicon cavity effort, we now achieve the fundamental thermal noise-limited stability for a 6 cm long silicon cavity cooled to 4 Kelvin, reaching 6.5 × 10 −17 from 0.8 to 80 seconds. We also report for the first time a clear linear dependence of the cavity frequency drift on the incident optical power. The lowest fractional frequency drift of −3 × 10 −19 /s is attained at a transmitted power of 40 nW, with an extrapolated drift approaching zero in the absence of optical power. These demonstrations provide a promising direction to reach a new performance domain for stable lasers, with stability better than 1 × 10 −17 and fractional linear drift below 1 ×

No abstract

Capturing associations for knowledge graphs (KGs) through entity alignment, entity type inference and other related tasks benefits NLP applications with comprehensive knowledge representations. Recent related methods built on Euclidean embeddings are challenged by the hierarchical structures and different scales of KGs. They also depend on high embedding dimensions to realize enough expressiveness. Differently, we explore with low-dimensional hyperbolic embeddings for knowledge association. We propose a hyperbolic relational graph neural network for KG embedding and capture knowledge associations with a hyperbolic transformation. Extensive experiments on entity alignment and type inference demonstrate the effectiveness and efficiency of our method.

No abstract

No abstract

No abstract

Nymphs of Aphis glycines Matsumura were individually reared to adults in the laboratory on detached leaf discs of Trifolium repens L. (white clover) mounted on agar medium. Adults of A. glycines were fed T. repens within small clip cages in the field. Development, reproduction and intrinsic rates of increase of A. glycines were studied. These data were compared to those of controls fed known host plants including cultivated soybean Glycine max (L.) Merr. and the wild soybean species Glycine soja Sieb & Zucc. The results demonstrated that nymphs of A. glycines successfully developed into adults and reproduced efficiently when reared on T. repens in the laboratory. The lower development temperature threshold for nymphs fed T. repens was estimated as 8.27 °C, and the effective cumulative temperature for A. glycines development from nymph to adult was 90.91 degree-days. Adults of A. glycines could also survive on T. repens in the field, but only a few nymphs were produced.