Generating images from textual descriptions has gained a lot of attention. Recently, DALL-E, a multimodal transformer language model, and its variants have shown high-quality text-to-image generation capabilities with a simple architecture and training objective, powered by large-scale training data and computation. However, despite the interesting image generation results, there has not been a detailed analysis on how to evaluate such models. In this work, we investigate the reasoning capabilities and social biases of such text-to-image generative transformers in detail. First, we measure four visual reasoning skills: object recognition, object counting, color recognition, and spatial relation understanding. For this, we propose PaintSkills, a diagnostic dataset and evaluation toolkit that measures these four visual reasoning skills. Second, we measure the text alignment and quality of the generated images based on pretrained image captioning, image-text retrieval, and image classification models. Third, we assess social biases in the models. For this, we suggest evaluation of gender and racial biases of text-to-image generation models based on a pretrained image-text retrieval model and human evaluation. In our experiments, we show that recent text-to-image models perform better in recognizing and counting objects than recognizing colors and understanding spatial relations, while there exists a large gap between model performances and oracle accuracy on all skills. Next, we demonstrate that recent text-to-image models learn specific gender/racial biases from web image-text pairs. We also show that our automatic evaluations of visual reasoning skills and gender bias are highly correlated with human judgments. We hope our work will help guide future progress in improving text-to-image models on visual reasoning skills and social biases. Code and data at: https://github.com/j-min/DallEval

Machine Learning has attracted considerable attention throughout the past decade due to its potential to solve far-reaching tasks, such as image classification, object recognition, anomaly detection, and data forecasting. A standard approach to tackle such applications is based on supervised learning, which is assisted by large sets of labeled data and is conducted by the so-called classifiers, such as Logistic Regression, Decision Trees, Random Forests, and Support Vector Machines, among others. An alternative to traditional classifiers is the parameterless Optimum-Path Forest (OPF), which uses a graph-based methodology and a distance measure to create arcs between nodes and hence sets of trees, responsible for conquering the nodes, defining their labels, and shaping the forests. Nevertheless, its performance is strongly associated with an appropriate distance measure, which may vary according to the dataset's nature. Therefore, this work proposes a comparative study over a wide range of distance measures applied to the supervised Optimum-Path Forest classification. The experimental results are conducted using well-known literature datasets and compared across benchmarking classifiers, illustrating OPF's ability to adapt to distinct domains.

In presence of multiple objectives to be optimized in Search-Based Software Engineering (SBSE), Pareto search has been commonly adopted. It searches for a good approximation of the problem's Pareto optimal solutions, from which the stakeholders choose the most preferred solution according to their preferences. However, when clear preferences of the stakeholders (e.g., a set of weights which reflect relative importance between objectives) are available prior to the search, weighted search is believed to be the first choice since it simplifies the search via converting the original multi-objective problem into a single-objective one and enable the search to focus on what only the stakeholders are interested in. This paper questions such a "weighted search first" belief. We show that the weights can, in fact, be harmful to the search process even in the presence of clear preferences. Specifically, we conduct a large scale empirical study which consists of 38 systems/projects from three representative SBSE problems, together with two types of search budget and nine sets of weights, leading to 604 cases of comparisons. Our key finding is that weighted search reaches a certain level of solution quality by consuming relatively less resources at the early stage of the search; however, Pareto search is at the majority of the time (up to 77% of the cases) significantly better than its weighted counterpart, as long as we allow a sufficient, but not unrealistic search budget. This, together with other findings and actionable suggestions in the paper, allows us to codify pragmatic and comprehensive guidance on choosing weighted and Pareto search for SBSE under the circumstance that clear preferences are available. All code and data can be accessed at: https://github.com/ideas-labo/pareto-vs-weight-for-sbse.

Learning representations of multimodal data that are both informative and robust to missing modalities at test time remains a challenging problem due to the inherent heterogeneity of data obtained from different channels. To address it, we present a novel Geometric Multimodal Contrastive (GMC) representation learning method comprised of two main components: i) a two-level architecture consisting of modality-specific base encoder, allowing to process an arbitrary number of modalities to an intermediate representation of fixed dimensionality, and a shared projection head, mapping the intermediate representations to a latent representation space; ii) a multimodal contrastive loss function that encourages the geometric alignment of the learned representations. We experimentally demonstrate that GMC representations are semantically rich and achieve state-of-the-art performance with missing modality information on three different learning problems including prediction and reinforcement learning tasks.

Recent progress in deep learning has relied on access to large and diverse datasets. Such data-driven progress has been less evident in offline reinforcement learning (RL), because offline RL data is usually collected to optimize specific target tasks limiting the data's diversity. In this work, we propose Exploratory data for Offline RL (ExORL), a data-centric approach to offline RL. ExORL first generates data with unsupervised reward-free exploration, then relabels this data with a downstream reward before training a policy with offline RL. We find that exploratory data allows vanilla off-policy RL algorithms, without any offline-specific modifications, to outperform or match state-of-the-art offline RL algorithms on downstream tasks. Our findings suggest that data generation is as important as algorithmic advances for offline RL and hence requires careful consideration from the community. Code and data can be found at https://github.com/denisyarats/exorl .

Multi-model ensemble analysis integrates information from multiple climate models into a unified projection. However, existing integration approaches based on model averaging can dilute fine-scale spatial information and incur bias from rescaling low-resolution climate models. We propose a statistical approach, called NN-GPR, using Gaussian process regression (GPR) with an infinitely wide deep neural network based covariance function. NN-GPR requires no assumptions about the relationships between models, no interpolation to a common grid, no stationarity assumptions, and automatically downscales as part of its prediction algorithm. Model experiments show that NN-GPR can be highly skillful at surface temperature and precipitation forecasting by preserving geospatial signals at multiple scales and capturing inter-annual variability. Our projections particularly show improved accuracy and uncertainty quantification skill in regions of high variability, which allows us to cheaply assess tail behavior at a 0.44$^\circ$/50 km spatial resolution without a regional climate model (RCM). Evaluations on reanalysis data and SSP245 forced climate models show that NN-GPR produces similar, overall climatologies to the model ensemble while better capturing fine scale spatial patterns. Finally, we compare NN-GPR's regional predictions against two RCMs and show that NN-GPR can rival the performance of RCMs using only global model data as input.

When is it a bad idea to bet on the future? First, we're seeing massive shifts in customer behavior during the pandemic -- behaviors that look likely to last. Second, the emergence of Millennials and Gen Z as significant market segments suggest that those new behaviors are just the beginning. Third, and most importantly, the big guys of digital -- Apple, Facebook, Google, Amazon, and Microsoft -- are all placing big bets that threaten to reshape the landscape for every business in due time. So, maybe a better question is "How can you bet on the future to win?" We'll take a look at who's leading the way towards the future, some useful frameworks for how to think about betting on the future, and how to place smart bets for your business… bets that you can win. Here are the show notes for you. Here are the regular show notes detailing links and news related to this week's episode.

Reusing previously trained models is critical in deep reinforcement learning to speed up training of new agents. However, it is unclear how to acquire new skills when objectives and constraints are in conflict with previously learned skills. Moreover, when retraining, there is an intrinsic conflict between exploiting what has already been learned and exploring new skills. In soft actor-critic (SAC) methods, a temperature parameter can be dynamically adjusted to weight the action entropy and balance the explore $\times$ exploit trade-off. However, controlling a single coefficient can be challenging within the context of retraining, even more so when goals are contradictory. In this work, inspired by neuroscience research, we propose a novel approach using inhibitory networks to allow separate and adaptive state value evaluations, as well as distinct automatic entropy tuning. Ultimately, our approach allows for controlling inhibition to handle conflict between exploiting less risky, acquired behaviors and exploring novel ones to overcome more challenging tasks. We validate our method through experiments in OpenAI Gym environments.

Language Models (LMs) often cannot be deployed because of their potential to harm users in hard-to-predict ways. Prior work identifies harmful behaviors before deployment by using human annotators to hand-write test cases. However, human annotation is expensive, limiting the number and diversity of test cases. In this work, we automatically find cases where a target LM behaves in a harmful way, by generating test cases ("red teaming") using another LM. We evaluate the target LM's replies to generated test questions using a classifier trained to detect offensive content, uncovering tens of thousands of offensive replies in a 280B parameter LM chatbot. We explore several methods, from zero-shot generation to reinforcement learning, for generating test cases with varying levels of diversity and difficulty. Furthermore, we use prompt engineering to control LM-generated test cases to uncover a variety of other harms, automatically finding groups of people that the chatbot discusses in offensive ways, personal and hospital phone numbers generated as the chatbot's own contact info, leakage of private training data in generated text, and harms that occur over the course of a conversation. Overall, LM-based red teaming is one promising tool (among many needed) for finding and fixing diverse, undesirable LM behaviors before impacting users.

To certify UAV operations in populated areas, risk mitigation strategies -- such as Emergency Landing (EL) -- must be in place to account for potential failures. EL aims at reducing ground risk by finding safe landing areas using on-board sensors. The first contribution of this paper is to present a new EL approach, in line with safety requirements introduced in recent research. In particular, the proposed EL pipeline includes mechanisms to monitor learning based components during execution. This way, another contribution is to study the behavior of Machine Learning Runtime Monitoring (MLRM) approaches within the context of a real-world critical system. A new evaluation methodology is introduced, and applied to assess the practical safety benefits of three MLRM mechanisms. The proposed approach is compared to a default mitigation strategy (open a parachute when a failure is detected), and appears to be much safer.