AI News, BOOK REVIEW: Machine Intelligence from Cortical Networks (MICrONS)

Machine Intelligence from Cortical Networks (MICrONS)

Participants in the program will have the unique opportunity to pose biological questions with the greatest potential to advance theories of neural computation and obtain answers through carefully planned experimentation and data analysis.

Over the course of the program, participants will use their improving understanding of the representations, transformations, and learning rules employed by the brain to create ever more capable neurally derived machine learning algorithms.

Machine Intelligence from Cortical Networks (MICrONS)

Participants in the program will have the unique opportunity to pose biological questions with the greatest potential to advance theories of neural computation and obtain answers through carefully planned experimentation and data analysis.

Over the course of the program, participants will use their improving understanding of the representations, transformations, and learning rules employed by the brain to create ever more capable neurally derived machine learning algorithms.

Mapping the Brain to Build Better Machines

“We are still more flexible in thinking and can anticipate, imagine and create future events.” An ambitious new program, funded by the federal government’s intelligence arm, aims to bring artificial intelligence more in line with our own mental powers.

Koch and his colleagues are now creating a complete wiring diagram of a small cube of brain — a million cubic microns, totaling one five-hundredth the volume of a poppy seed.

That tiny portion houses about 100,000 neurons, 3 to 15 million neuronal connections, or synapses, and enough neural wiring to span the width of Manhattan, were it all untangled and laid end-to-end.

In a paper published in the journal Nature in March, Wei-Chung Allen Lee — a neuroscientist at Harvard University who is working with Koch’s team — and his collaborators mapped out a wiring diagram of 50 neurons and more than 1,000 of their partners.

By pairing this map with information about each neuron’s job in the brain — some respond to a visual input of vertical bars, for example — they derived a simple rule for how neurons in this part of the cortex are anatomically connected.

While the implicit goal of the Microns project is technological — IARPA funds research that could eventually lead to data-analysis tools for the intelligence community, among other things — new and profound insights into the brain will have to come first.

Without knowing all the component parts, he said, “maybe we’re missing the beauty of the structure.” The Brain’s Processing Units The convoluted folds covering the brain’s surface form the cerebral cortex, a pizza-sized sheet of tissue that’s scrunched to fit into our skulls.

New technologies designed to trace the shape, activity and connectivity of thousands of neurons are finally allowing researchers to analyze how cells within a module interact with each other;

“Different teams have different guesses for what’s inside.” The researchers will focus on a part of the cortex that processes vision, a sensory system that neuroscientists have explored intensively and that computer scientists have long striven to emulate.

Tai Sing Lee’s team, co-led by George Church, theorizes that the brain has built a library of parts — bits and pieces of objects and people — and learns rules for how to put those parts together.

The team will then computationally stitch together each cross section to create a densely packed three-dimensional map that charts millions of neural wires on their intricate path through the cortex.

“That’s what [Tolias] has started to do.” Among these thousands of neuronal connections, Tolias’s team uncovered three general rules that govern how the cells are connected: Some talk mainly to neurons of their own kind;

(Tolias’s team defined their cells based on neural anatomy rather than function, which Wei Lee’s team did in their study.) Using just these three wiring rules, the researchers could simulate the circuit fairly accurately.

And although neural networks have enjoyed a major renaissance — the voice- and face-recognition programs that have rapidly become part of our daily lives are based on neural network algorithms, as is AlphaGo, the computer that recently defeated the world’s top Go player — the rules that artificial neural networks use to alter their connections are almost certainly different than the ones employed by the brain.

Contemporary neural networks “are based on what we knew about the brain in the 1960s,” said Terry Sejnowski, a computational neuroscientist at the Salk Institute in San Diego who developed early neural network algorithms with Geoffrey Hinton, a computer scientist at the University of Toronto.

“Our knowledge of how the brain is organized is exploding.” For example, today’s neural networks are comprised of a feed-forward architecture, where information flows from input to output through a series of layers.

Microns researchers aim to decipher the rules governing feedback loops — such as which cells these loops connect, what triggers their activity, and how that activity effects the circuit’s output — then translate those rules into an algorithm.

“If you could implement [feedback circuitry] in a deep network, you could go from a network that has kind of a knee-jerk reaction — give input and get output — to one that’s more reflective, that can start thinking about inputs and testing hypotheses,” said Sejnowski, who serves as an advisor to President Obama’s $100 million BRAIN Initiative, of which the Microns project is a part.

One challenge will be dealing with the enormous amounts of data the research produces — 1 to 2 petabytes of data per millimeter cube of brain.

Making The Connection

For six painstaking years, Jeff Lichtman and his team at Harvard University worked to assemble a complete 3-D map of a bit of brain so tiny it is about the breadth of a human hair.

Finished in 2015, that comprehensive reconstruction represented the largest portion of mammalian brain ever rendered in full detail — it measured 1,500 cubic microns and documented every cell type and connection.

Developing wiring diagrams ranging in scale from the individual synapse like Lichtman’s to the whole brain is a high priority for the BRAIN Initiative — a research effort launched by President Barack Obama in April 2013 to catalyze and speed advances in neuroscience by deploying cutting-edge computer science, physics, biology, and chemistry to develop transformative new tools.

Understanding how the brain processes information at the level of the circuit in addition to the physical structure of the connections could provide insights into how our brains behave when they are healthy, and what goes wrong in disorders like autism and schizophrenia when those circuits fail to hook up properly.

“Previous imaging techniques were just too limiting.” Lichtman will use that same process combined with an even more powerful microscope in a project sponsored by the Intelligence Advanced Research Projects Activity (IARPA) aimed at quantifying why brains are so good at learning in an effort to improve the algorithms for machine learning and artificial intelligence.

Neural network algorithms are already sophisticated enough to beat humans at the complex Chinese strategy game Go, and are learning how to drive automobiles, but computers still can’t match human intelligence in more complex settings such as understanding language.

Faced with the realization that a single person trying to trace the circuits in a cubic millimeter of human cortex would need nearly million years to complete the task, Seung decided to engage citizen scientists to help map the mouse retina.

In 2014, Seung published a paper detailing a wiring diagram for a circuit including starburst amacrine cells — a type of neuron with dendrites extending in all directions that has been implicated in detecting moving objects.

“The microscopic data is of unprecedented scale and detail.” Reid will use this combination of methods to tackle IARPA’s challenge to reverse-engineer one cubic millimeter of the mouse brain by recording the activity and connectivity of 100,000 neurons of the animal that had completed visual perception and learning tasks.

However, Lichtman notes the information generated will also provide valuable insights into how our brains make us human as scientists seek “those regularities that will allow us to understand how … information about the world gets physically built into your brain to create memories, personality traits and skills that you carry with you throughout your life.”

The Emerging World of ECoG Neuroprosthetics

Google Tech Talk September 27, 2010 Presented by Dr. Eric C. Leuthardt, Washington University. The notion that a computer can decode brain signals to infer the intentions of a human and then...

The World We Dream- Lisa Randall & Ron Garan Zeitgeist Americas 2012

The World We Dream-- Lisa Randall, Professor of Physics, Harvard University; Ron Garan, NASA Astronaut. Putting a man on the moon was once simply a dream. What are some of today's most innovative...

Developing neuroprosthetic treatments for spinal cord injury

Visit our website at Dr. Chet Moritz conducts research in brain-computer interfaces and neuroprosthetic technology. The goal of his research is to develop systems..

Reproducibility of Data Collection and Analysis

Reproducibility of Data Collection and Analysis -- Modern technologies in genome technology: potentials and pitfalls Air date: Thursday, June 04, 2015, 12:30:00 PM Category: Conferences...

Bode lecture CDC 2016 Richard Murray

Taken from The video is uploaded here in order to interlay subtitle for translation

New Fundamental Technologies by Dave Wilson

SPARC Strategic Planning Workshop: Biology & Technology (Day 2)

SPARC Strategic Planning Workshop: Biology & Technology (Day 2) Air date: Thursday, February 26, 2015, 8:00:00 AM Category: Advisory Board Meetings Runtime: 04:17:09 Description: The...