Large-scale diffusion neural networks represent a substantial milestone in text-to-image generation, but they remain poorly understood, lacking interpretability analyses. In this paper, we perform a text-image attribution analysis on Stable Diffusion, a recently open-sourced model. To produce pixel-level attribution maps, we upscale and aggregate cross-attention word-pixel scores in the denoising subnetwork, naming our method DAAM. We evaluate its correctness by testing its semantic segmentation ability on nouns, as well as its generalized attribution quality on all parts of speech, rated by humans. We then apply DAAM to study the role of syntax in the pixel space, characterizing head--dependent heat map interaction patterns for ten common dependency relations. Finally, we study several semantic phenomena using DAAM, with a focus on feature entanglement, where we find that cohyponyms worsen generation quality and descriptive adjectives attend too broadly. To our knowledge, we are the first to interpret large diffusion models from a visuolinguistic perspective, which enables future lines of research. Our code is at https://github.com/castorini/daam.

For training neural networks, flat-minima optimizers that seek to find parameters in neighborhoods having uniformly low loss (flat minima) have been shown to improve upon stochastic and adaptive gradient-based methods. Two methods for finding flat minima stand out: 1. Averaging methods (i.e., Stochastic Weight Averaging, SWA), and 2. Minimax methods (i.e., Sharpness Aware Minimization, SAM). However, despite similar motivations, there has been limited investigation into their properties and no comprehensive comparison between them. In this work, we investigate the loss surfaces from a systematic benchmarking of these approaches across computer vision, natural language processing, and graph learning tasks. The results lead to a simple hypothesis: since both approaches find different flat solutions, combining them should improve generalization even further. We verify this improves over either flat-minima approach in 39 out of 42 cases. When it does not, we investigate potential reasons. We hope our results across image, graph, and text data will help researchers to improve deep learning optimizers, and practitioners to pinpoint the optimizer for the problem at hand.

Recent work on Open Domain Question Answering has shown that there is a large discrepancy in model performance between novel test questions and those that largely overlap with training questions. However, it is as of yet unclear which aspects of novel questions that make them challenging. Drawing upon studies on systematic generalization, we introduce and annotate questions according to three categories that measure different levels and kinds of generalization: training set overlap, compositional generalization (comp-gen), and novel entity generalization (novel-entity). When evaluating six popular parametric and non-parametric models, we find that for the established Natural Questions and TriviaQA datasets, even the strongest model performance for comp-gen/novel-entity is 13.1/5.4% and 9.6/1.5% lower compared to that for the full test set -- indicating the challenge posed by these types of questions. Furthermore, we show that whilst non-parametric models can handle questions containing novel entities, they struggle with those requiring compositional generalization. Through thorough analysis we find that key question difficulty factors are: cascading errors from the retrieval component, frequency of question pattern, and frequency of the entity.

Open-domain Question Answering models which directly leverage question-answer (QA) pairs, such as closed-book QA (CBQA) models and QA-pair retrievers, show promise in terms of speed and memory compared to conventional models which retrieve and read from text corpora. QA-pair retrievers also offer interpretable answers, a high degree of control, and are trivial to update at test time with new knowledge. However, these models lack the accuracy of retrieve-and-read systems, as substantially less knowledge is covered by the available QA-pairs relative to text corpora like Wikipedia. To facilitate improved QA-pair models, we introduce Probably Asked Questions (PAQ), a very large resource of 65M automatically-generated QA-pairs. We introduce a new QA-pair retriever, RePAQ, to complement PAQ. We find that PAQ preempts and caches test questions, enabling RePAQ to match the accuracy of recent retrieve-and-read models, whilst being significantly faster. Using PAQ, we train CBQA models which outperform comparable baselines by 5%, but trail RePAQ by over 15%, indicating the effectiveness of explicit retrieval. RePAQ can be configured for size (under 500MB) or speed (over 1K questions per second) whilst retaining high accuracy. Lastly, we demonstrate RePAQ's strength at selective QA, abstaining from answering when it is likely to be incorrect. This enables RePAQ to ``back-off" to a more expensive state-of-the-art model, leading to a combined system which is both more accurate and 2x faster than the state-of-the-art model alone.

We review the EfficientQA competition from NeurIPS 2020. The competition focused on open-domain question answering (QA), where systems take natural language questions as input and return natural language answers. The aim of the competition was to build systems that can predict correct answers while also satisfying strict on-disk memory budgets. These memory budgets were designed to encourage contestants to explore the trade-off between storing large, redundant, retrieval corpora or the parameters of large learned models. In this report, we describe the motivation and organization of the competition, review the best submissions, and analyze system predictions to inform a discussion of evaluation for open-domain QA.

In this paper, we propose an adversarial process for abstractive text summarization, in which we simultaneously train a generative model G and a discriminative model D. In particular, we build the generator G as an agent of reinforcement learning, which takes the raw text as input and predicts the abstractive summarization. We also build a discriminator which attempts to distinguish the generated summary from the ground truth summary. Extensive experiments demonstrate that our model achieves competitive ROUGE scores with the state-of-the-art methods on CNN/Daily Mail dataset. Qualitatively, we show that our model is able to generate more abstractive, readable and diverse summaries.