Collaborating Authors


Dissonance Between Human and Machine Understanding Artificial Intelligence

Complex machine learning models are deployed in several critical domains including healthcare and autonomous vehicles nowadays, albeit as functional black boxes. Consequently, there has been a recent surge in interpreting decisions of such complex models in order to explain their actions to humans. Models that correspond to human interpretation of a task are more desirable in certain contexts and can help attribute liability, build trust, expose biases and in turn build better models. It is, therefore, crucial to understand how and which models conform to human understanding of tasks. In this paper, we present a large-scale crowdsourcing study that reveals and quantifies the dissonance between human and machine understanding, through the lens of an image classification task. In particular, we seek to answer the following questions: Which (well-performing) complex ML models are closer to humans in their use of features to make accurate predictions? How does task difficulty affect the feature selection capability of machines in comparison to humans? Are humans consistently better at selecting features that make image recognition more accurate? Our findings have important implications on human-machine collaboration, considering that a long term goal in the field of artificial intelligence is to make machines capable of learning and reasoning like humans.

Memetics and Neural Models of Conspiracy Theories Artificial Intelligence

Conspiracy theories, or in general seriously distorted beliefs, are widespread. How and why are they formed in the brain is still more a matter of speculation rather than science. In this paper one plausible mechanisms is investigated: rapid freezing of high neuroplasticity (RFHN). Emotional arousal increases neuroplasticity and leads to creation of new pathways spreading neural activation. Using the language of neurodynamics a meme is defined as quasi-stable associative memory attractor state. Depending on the temporal characteristics of the incoming information and the plasticity of the network, memory may self-organize creating memes with large attractor basins, linking many unrelated input patterns. Memes with fake rich associations distort relations between memory states. Simulations of various neural network models trained with competitive Hebbian learning (CHL) on stationary and non-stationary data lead to the same conclusion: short learning with high plasticity followed by rapid decrease of plasticity leads to memes with large attraction basins, distorting input pattern representations in associative memory. Such system-level models may be used to understand creation of distorted beliefs and formation of conspiracy memes, understood as strong attractor states of the neurodynamics.

Faster Convergence in Deep-Predictive-Coding Networks to Learn Deeper Representations Artificial Intelligence

Deep-predictive-coding networks (DPCNs) are hierarchical, generative models that rely on feed-forward and feed-back connections to modulate latent feature representations of stimuli in a dynamic and context-sensitive manner. A crucial element of DPCNs is a forward-backward inference procedure to uncover sparse states of a dynamic model, which are used for invariant feature extraction. However, this inference and the corresponding backwards network parameter updating are major computational bottlenecks. They severely limit the network depths that can be reasonably implemented and easily trained. We therefore propose a optimization strategy, with better empirical and theoretical convergence, based on accelerated proximal gradients. We demonstrate that the ability to construct deeper DPCNs leads to receptive fields that capture well the entire notions of objects on which the networks are trained. This improves the feature representations. It yields completely unsupervised classifiers that surpass convolutional and convolutional-recurrent autoencoders and are on par with convolutional networks trained in a supervised manner. This is despite the DPCNs having orders of magnitude fewer parameters.

Neurocognitive Informatics Manifesto Artificial Intelligence

Theoretical and abstract approaches to information have made great advances, but human information processing is still unmatched in many areas, including information management, representation and understanding. Neurocognitive informatics is a new, emerging field that should help to improve the matching of artificial and natural systems, and inspire better computational algorithms to solve problems that are still beyond the reach of machines. In this position paper examples of neurocognitive inspirations and promising directions in this area are given.

Slow manifolds in recurrent networks encode working memory efficiently and robustly Artificial Intelligence

Working memory is a cognitive function involving the storage and manipulation of latent information over brief intervals of time, thus making it crucial for context-dependent computation. Here, we use a top-down modeling approach to examine network-level mechanisms of working memory, an enigmatic issue and central topic of study in neuroscience and machine intelligence. We train thousands of recurrent neural networks on a working memory task and then perform dynamical systems analysis on the ensuing optimized networks, wherein we find that four distinct dynamical mechanisms can emerge. In particular, we show the prevalence of a mechanism in which memories are encoded along slow stable manifolds in the network state space, leading to a phasic neuronal activation profile during memory periods. In contrast to mechanisms in which memories are directly encoded at stable attractors, these networks naturally forget stimuli over time. Despite this seeming functional disadvantage, they are more efficient in terms of how they leverage their attractor landscape and paradoxically, are considerably more robust to noise. Our results provide new dynamical hypotheses regarding how working memory function is encoded in both natural and artificial neural networks.

On the Control of Attentional Processes in Vision Artificial Intelligence

The study of attentional processing in vision has a long and deep history. Recently, several papers have presented insightful perspectives into how the coordination of multiple attentional functions in the brain might occur. These begin with experimental observations and the authors propose structures, processes, and computations that might explain those observations. Here, we consider a perspective that past works have not, as a complementary approach to the experimentally-grounded ones. We approach the same problem as past authors but from the other end of the computational spectrum, from the problem nature, as Marr's Computational Level would prescribe. What problem must the brain solve when orchestrating attentional processes in order to successfully complete one of the myriad possible visuospatial tasks at which we as humans excel? The hope, of course, is for the approaches to eventually meet and thus form a complete theory, but this is likely not soon. We make the first steps towards this by addressing the necessity of attentional control, examining the breadth and computational difficulty of the visuospatial and attentional tasks seen in human behavior, and suggesting a sketch of how attentional control might arise in the brain. The key conclusions of this paper are that an executive controller is necessary for human attentional function in vision, and that there is a 'first principles' computational approach to its understanding that is complementary to the previous approaches that focus on modelling or learning from experimental observations directly.

An Existential Crisis in Neuroscience - Issue 94: Evolving


This week we are reprinting our top stories of 2020. This article first appeared online in our "Maps" issue in January, 2020. On a chilly evening last fall, I stared into nothingness out of the floor-to-ceiling windows in my office on the outskirts of Harvard's campus. As a purplish-red sun set, I sat brooding over my dataset on rat brains. I thought of the cold windowless rooms in downtown Boston, home to Harvard's high-performance computing center, where computer servers were holding on to a precious 48 terabytes of my data. I have recorded the 13 trillion numbers in this dataset as part of my Ph.D. experiments, asking how the visual parts of the rat brain respond to movement. Printed on paper, the dataset would fill 116 billion pages, double-spaced. When I recently finished writing the story of my data, the magnum opus fit on fewer than two dozen printed pages. Performing the experiments turned out to be the easy part.

Top 100 Artificial Intelligence Companies in the World


Artificial Intelligence (AI) is not just a buzzword, but a crucial part of the technology landscape. AI is changing every industry and business function, which results in increased interest in its applications, subdomains and related fields. This makes AI companies the top leaders driving the technology swift. AI helps us to optimise and automate crucial business processes, gather essential data and transform the world, one step at a time. From Google and Amazon to Apple and Microsoft, every major tech company is dedicating resources to breakthroughs in artificial intelligence. As big enterprises are busy acquiring or merging with other emerging inventions, small AI companies are also working hard to develop their own intelligent technology and services. By leveraging artificial intelligence, organizations get an innovative edge in the digital age. AI consults are also working to provide companies with expertise that can help them grow. In this digital era, AI is also a significant place for investment. AI companies are constantly developing the latest products to provide the simplest solutions. Henceforth, Analytics Insight brings you the list of top 100 AI companies that are leading the technology drive towards a better tomorrow. AEye develops advanced vision hardware, software, and algorithms that act as the eyes and visual cortex of autonomous vehicles. AEye is an artificial perception pioneer and creator of iDAR, a new form of intelligent data collection that acts as the eyes and visual cortex of autonomous vehicles. Since its demonstration of its solid state LiDAR scanner in 2013, AEye has pioneered breakthroughs in intelligent sensing. Their mission was to acquire the most information with the fewest ones and zeros. This would allow AEye to drive the automotive industry into the next realm of autonomy. Algorithmia invented the AI Layer.

Insights for AI from the Human Mind

Communications of the ACM

What magical trick makes us intelligent? The trick is that there is no trick. The power of intelligence stems from our vast diversity, not from any single, perfect principle. Artificial intelligence has recently beaten world champions in Go and poker and made extraordinary progress in domains such as machine translation, object classification, and speech recognition. However, most AI systems are extremely narrowly focused. AlphaGo, the champion Go player, does not know that the game is played by putting stones onto a board; it has no idea what a "stone" or a "board" is, and would need to be retrained from scratch if you presented it with a rectangular board rather than a square grid.

Applying Deutsch's concept of good explanations to artificial intelligence and neuroscience -- an initial exploration Artificial Intelligence

Artificial intelligence has made great strides since the deep learning revolution, but AI systems still struggle to extrapolate outside of their training data and adapt to new situations. For inspiration we look to the domain of science, where scientists have been able to develop theories which show remarkable ability to extrapolate and sometimes predict the existence of phenomena which have never been observed before. According to David Deutsch, this type of extrapolation, which he calls "reach", is due to scientific theories being hard to vary. In this work we investigate Deutsch's hard-to-vary principle and how it relates to more formalized principles in deep learning such as the bias-variance trade-off and Occam's razor. We distinguish internal variability, how much a model/theory can be varied internally while still yielding the same predictions, with external variability, which is how much a model must be varied to accurately predict new, out-of-distribution data. We discuss how to measure internal variability using the size of the Rashomon set and how to measure external variability using Kolmogorov complexity. We explore what role hard-to-vary explanations play in intelligence by looking at the human brain and distinguish two learning systems in the brain. The first system operates similar to deep learning and likely underlies most of perception and motor control while the second is a more creative system capable of generating hard-to-vary explanations of the world. We argue that figuring out how replicate this second system, which is capable of generating hard-to-vary explanations, is a key challenge which needs to be solved in order to realize artificial general intelligence. We make contact with the framework of Popperian epistemology which rejects induction and asserts that knowledge generation is an evolutionary process which proceeds through conjecture and refutation.