Explore the vast range of titles available.

Humans readily form social impressions, such as attractiveness and trustworthiness, from a stranger’s facial features. Understanding the provenance of these impressions has clear scientific importance and societal implications. Motivated by the efficient coding hypothesis of brain representation, as well as Claude Shannon’s theoretical result that maximally efficient representational systems assign shorter codes to statistically more typical data (quantified as log likelihood), we suggest that social “liking” of faces increases with statistical typicality. Combining human behavioral data and computational modeling, we show that perceived attractiveness, trustworthiness, dominance, and valence of a face image linearly increase with its statistical typicality (log likelihood). We also show that statistical typicality can at least partially explain the role of symmetry in attractiveness perception. Additionally, by assuming that the brain focuses on a task-relevant subset of facial features and assessing log likelihood of a face using those features, our model can explain the “ugliness-in-averageness” effect found in social psychology, whereby otherwise attractive, intercategory faces diminish in attractiveness during a categorization task.

Understanding how humans and animals learn about statistical regularities in stable and volatile environments, and utilize these regularities to make predictions and decisions, is an important problem in neuroscience and psychology. Using a Bayesian modeling framework, specifically the Dynamic Belief Model (DBM), it has previously been shown that humans tend to make the default assumption that environmental statistics undergo abrupt, unsignaled changes, even when environmental statistics are actually stable. Because exact Bayesian inference in this setting, an example of switching state space models, is computationally intensive, a number of approximately Bayesian and heuristic algorithms have been proposed to account for learning/prediction in the brain. Here, we examine a neurally plausible algorithm, a special case of leaky integration dynamics we denote as EXP (for exponential filtering), that is significantly simpler than all previously suggested algorithms except for the delta-learning rule, and which far outperforms the delta rule in approximating Bayesian prediction performance. We derive the theoretical relationship between DBM and EXP, and show that EXP gains computational efficiency by foregoing the representation of inferential uncertainty (as does the delta rule), but that it nevertheless achieves near-Bayesian performance due to its ability to incorporate a \"persistent prior\" influence unique to DBM and absent from the other algorithms. Furthermore, we show that EXP is comparable to DBM but better than all other models in reproducing human behavior in a visual search task, suggesting that human learning and prediction also incorporates an element of persistent prior. More broadly, our work demonstrates that when observations are information-poor, detecting changes or modulating the learning rate is both difficult and (thus) unnecessary for making Bayes-optimal predictions.

Recent progress in self-supervised learning has demonstrated promising results in multiple visual tasks. An important ingredient in high-performing self-supervised methods is the use of data augmentation by training models to place different augmented views of the same image nearby in embedding space. However, commonly used augmentation pipelines treat images holistically, ignoring the semantic relevance of parts of an image-e.g. a subject vs. a background-which can lead to the learning of spurious correlations. Our work addresses this problem by investigating a class of simple, yet highly effective \"background augmentations\", which encourage models to focus on semantically-relevant content by discouraging them from focusing on image backgrounds. Through a systematic investigation, we show that background augmentations lead to substantial improvements in performance across a spectrum of state-of-the-art self-supervised methods (MoCo-v2, BYOL, SwAV) on a variety of tasks, e.g. \\(\\sim\\)+1-2% gains on ImageNet, enabling performance on par with the supervised baseline. Further, we find the improvement in limited-labels settings is even larger (up to 4.2%). Background augmentations also improve robustness to a number of distribution shifts, including natural adversarial examples, ImageNet-9, adversarial attacks, ImageNet-Renditions. We also make progress in completely unsupervised saliency detection, in the process of generating saliency masks used for background augmentations.

The fruit fly Drosophila's olfactory circuit has inspired a new locality sensitive hashing (LSH) algorithm, FlyHash. In contrast with classical LSH algorithms that produce low dimensional hash codes, FlyHash produces sparse high-dimensional hash codes and has also been shown to have superior empirical performance compared to classical LSH algorithms in similarity search. However, FlyHash uses random projections and cannot learn from data. Building on inspiration from FlyHash and the ubiquity of sparse expansive representations in neurobiology, our work proposes a novel hashing algorithm BioHash that produces sparse high dimensional hash codes in a data-driven manner. We show that BioHash outperforms previously published benchmarks for various hashing methods. Since our learning algorithm is based on a local and biologically plausible synaptic plasticity rule, our work provides evidence for the proposal that LSH might be a computational reason for the abundance of sparse expansive motifs in a variety of biological systems. We also propose a convolutional variant BioConvHash that further improves performance. From the perspective of computer science, BioHash and BioConvHash are fast, scalable and yield compressed binary representations that are useful for similarity search.

The mushroom body of the fruit fly brain is one of the best studied systems in neuroscience. At its core it consists of a population of Kenyon cells, which receive inputs from multiple sensory modalities. These cells are inhibited by the anterior paired lateral neuron, thus creating a sparse high dimensional representation of the inputs. In this work we study a mathematical formalization of this network motif and apply it to learning the correlational structure between words and their context in a corpus of unstructured text, a common natural language processing (NLP) task. We show that this network can learn semantic representations of words and can generate both static and context-dependent word embeddings. Unlike conventional methods (e.g., BERT, GloVe) that use dense representations for word embedding, our algorithm encodes semantic meaning of words and their context in the form of sparse binary hash codes. The quality of the learned representations is evaluated on word similarity analysis, word-sense disambiguation, and document classification. It is shown that not only can the fruit fly network motif achieve performance comparable to existing methods in NLP, but, additionally, it uses only a fraction of the computational resources (shorter training time and smaller memory footprint).

Understanding how humans perceive the likability of high-dimensional ``objects'' such as faces is an important problem in both cognitive science and AI/ML. Existing models of human preferences generally assume these preferences to be fixed. However, human assessment of facial attractiveness have been found to be highly context-dependent. Specifically, the classical Beauty-in-Averageness (BiA) effect, whereby a face blended from two original faces is judged to be more attractive than the originals, is significantly diminished or reversed when the original faces are recognizable, or when the morph is mixed-race/mixed gender and the attractiveness judgment is preceded by a race/gender categorization. This effect, dubbed Ugliness-in-Averageness (UiA), has previously been attributed to a disfluency account, which is both qualitative and clumsy in explaining BiA. We hypothesize, instead, that these contextual influences on face processing result from the dependence of attractiveness perception on an element of statistical typicality, and from an attentional mechanism that restricts face representation to a task-relevant subset of features, thus redefining typicality within that subspace. Furthermore, we propose a principled explanation of why statistically atypical objects are less likable: they incur greater encoding or processing cost associated with a greater prediction error, when the brain uses predictive coding to compare the actual stimulus properties with those expected from its associated categorical prototype. We use simulations to show our model provides a parsimonious, statistically grounded, and quantitative account of contextual dependence of attractiveness. We also validate our model using experimental data from a gender categorization task. Finally, we make model predictions for a proposed experiment that can disambiguate the previous disfluency account and our statistical typicality theory.

Face processing plays a central role in human social life. Humans readily infer social traits (e.g. attractiveness and trustworthiness) from a stranger's face. Previous attempts to characterize the facial (physiognomic) features underlying social processing have lacked either systematicity or interpretability. Here, we utilize a statistical framework to tackle this problem, by learning a vector space to represent faces, and a linear mapping from this face space into human social trait judgments. Specifically, we obtain a face space by training the Active Appearance Model on large datasets of face images. Based on human evaluations of numerous social traits on these images, we then use regression to find linear combinations of facial features (what we call Linear Trait Axis, or LTA) that best predict human social judgments. Our model achieves state-of-the-art performance in overall predictive accuracy -- comparable to the best convolutional neural network and better than human prediction of other human ratings. To interpret the LTAs, we regress them against a large repertoire of geometric features. To understand the relationship between the facial features that underlie different social, emotional, and demographic traits, we present a novel \"dual space analysis\" that characterizes the geometric relationship among LTA vectors. It shows that facial features important for social trait perception are largely distinct from those underlying demographic and emotion perception, contrary to previous suggestions that social trait perception is driven by over-generalization of relatively primitive demographic and emotion perception processes. In addition, we present a novel correlation decomposition analysis that quantifies how correlations in trait judgments (e.g. between attractiveness and babyfacedness) independently arise from (1) shared facial features among traits, and (2) correlation in the distribution of facial features in the human population.