susannah young

This paper describes the DeepMind submission to the Chinese\rightarrowEnglish constrained data track of the WMT2020 Shared Task on News Translation. The submission employs a noisy channel factorization as the backbone of a document translation system. This approach allows the flexible combination of a number of independent component models which are further augmented with back-translation, distillation, fine-tuning with in-domain data, Monte-Carlo Tree Search decoding, and improved uncertainty estimation. In order to address persistent issues with the premature truncation of long sequences we included specialized length models and sentence segmentation techniques. Our final system provides a 9.9 BLEU points improvement over a baseline Transformer on our test set (newstest 2019).

In natural images, transitions between adjacent pixels tend to be smooth and gradual, a fact that has long been exploited in image compression models based on predictive coding. In contrast, existing neural autoregressive image generation models predict the absolute pixel intensities at each position, which is a more challenging problem. In this paper, we propose to predict pixels relatively, by predicting new pixels relative to previously generated pixels (or pixels from the conditioning context, when available). We show that this form of prediction fare favorably to its absolute counterpart when used independently, but their coordination under an unified probabilistic model yields optimal performance, as the model learns to predict sharp transitions using the absolute predictor, while generating smooth transitions using the relative predictor. Experiments on multiple benchmarks for unconditional image generation, image colorization, and super-resolution indicate that our presented...

Our world is open-ended, non-stationary and constantly evolving; thus what we talk about and how we talk about it changes over time. This inherent dynamic nature of language comes in stark contrast to the current static language modelling paradigm, which constructs training and evaluation sets from overlapping time periods. Despite recent progress, we demonstrate that state-of-the-art Transformer models perform worse in the realistic setup of predicting future utterances from beyond their training period—a consistent pattern across three datasets from two domains. We find that, while increasing model size alone—a key driver behind recent progress—does not provide a solution for the temporal generalization problem, having models that continually update their knowledge with new information can indeed slow down the degradation over time. Hence, given the compilation of ever-larger language modelling training datasets, combined with the growing list of language-model-based NLP applicati...

Our world is open-ended, non-stationary, and constantly evolving; thus what we talk about and how we talk about it change over time. This inherent dynamic nature of language contrasts with the current static language modelling paradigm, which trains and evaluates models on utterances from overlapping time periods. Despite impressive recent progress, we demonstrate that Transformer-XL language models perform worse in the realistic setup of predicting future utterances from beyond their training period, and that model performance becomes increasingly worse with time. We find that, while increasing model size alone—a key driver behind recent progress—does not solve this problem, having models that continually update their knowledge with new information can indeed mitigate this performance degradation over time. Hence, given the compilation of ever-larger language modelling datasets, combined with the growing list of language-model-based NLP applications that require up-to-date factual ...

Language modelling provides a step towards intelligent communication systems by harnessing large repositories of written human knowledge to better predict and understand the world. In this paper, we present an analysis of Transformer-based language model performance across a wide range of model scales -- from models with tens of millions of parameters up to a 280 billion parameter model called Gopher. These models are evaluated on 152 diverse tasks, achieving state-of-the-art performance across the majority. Gains from scale are largest in areas such as reading comprehension, fact-checking, and the identification of toxic language, but logical and mathematical reasoning see less benefit. We provide a holistic analysis of the training dataset and model's behaviour, covering the intersection of model scale with bias and toxicity. Finally we discuss the application of language models to AI safety and the mitigation of downstream harms.