Chapter 17. Learning and Memory

Follow us on Facebook and Twitter, or subscribe to our mailing list, to receive news updates. Learn more.


Links 1 - 20 of 1448

Carl Zimmer When you drive toward an intersection, the sight of the light turning red will (or should) make you step on the brake. This action happens thanks to a chain of events inside your head. Your eyes relay signals to the visual centers in the back of your brain. After those signals get processed, they travel along a pathway to another region, the premotor cortex, where the brain plans movements. Now, imagine that you had a device implanted in your brain that could shortcut the pathway and “inject” information straight into your premotor cortex. That may sound like an outtake from “The Matrix.” But now two neuroscientists at the University of Rochester say they have managed to introduce information directly into the premotor cortex of monkeys. The researchers published the results of the experiment on Thursday in the journal Neuron. Although the research is preliminary, carried out in just two monkeys, the researchers speculated that further research might lead to brain implants for people with strokes. “You could potentially bypass the damaged areas and deliver stimulation to the premotor cortex,” said Kevin A. Mazurek, a co-author of the study. “That could be a way to bridge parts of the brain that can no longer communicate.” In order to study the premotor cortex, Dr. Mazurek and his co-author, Dr. Marc H. Schieber, trained two rhesus monkeys to play a game. The monkeys sat in front of a panel equipped with a button, a sphere-shaped knob, a cylindrical knob, and a T-shaped handle. Each object was ringed by LED lights. If the lights around an object switched on, the monkeys had to reach out their hand to it to get a reward — in this case, a refreshing squirt of water. © 2017 The New York Times Company

Keyword: Learning & Memory; Movement Disorders
Link ID: 24408 - Posted: 12.08.2017

By David Z. Hambrick, Madeline Marquardt There are advantages to being smart. People who do well on standardized tests of intelligence—IQ tests—tend to be more successful in the classroom and the workplace. Although the reasons are not fully understood, they also tend to live longer, healthier lives, and are less likely to experience negative life events such as bankruptcy. Now there’s some bad news for people in the right tail of the IQ bell curve. In a study just published in the journal Intelligence, Pitzer College researcher Ruth Karpinski and her colleagues emailed a survey with questions about psychological and physiological disorders to members of Mensa. A “high IQ society”, Mensa requires that its members have an IQ in the top two percent. For most intelligence tests, this corresponds to an IQ of about 132 or higher. (The average IQ of the general population is 100.) The survey of Mensa’s highly intelligent members found that they were more likely to suffer from a range of serious disorders. The survey covered mood disorders (depression, dysthymia, and bipolar), anxiety disorders (generalized, social, and obsessive-compulsive), attention-deficit hyperactivity disorder, and autism. It also covered environmental allergies, asthma, and autoimmune disorders. Respondents were asked to report whether they had ever been formally diagnosed with each disorder, or suspected they suffered from it. With a return rate of nearly 75%, Karpinski and colleagues compared the percentage of the 3,715 respondents who reported each disorder to the national average. © 2017 Scientific American

Keyword: Intelligence; Depression
Link ID: 24397 - Posted: 12.06.2017

By Rebecca Robbins, Akili Interactive Labs on Monday reported that its late-stage study of a video game designed to treat kids with ADHD met its primary goal, a big step in the Boston company’s quest to get approval for what it hopes will be the first prescription video game. In a study of 348 children between the ages of 8 and 12 diagnosed with ADHD, those who played Akili’s action-packed game on a tablet over four weeks saw statistically significant improvements on metrics of attention and inhibitory control, compared to children who were given a different action-driven video game designed as a placebo. The company plans next year to file for approval with the Food and Drug Administration. “We are directly targeting the key neurological pathways that control attention and impulsivity,” said Akili CEO Eddie Martucci. The study “was meant to be a strong objective test to ask: Is it the targeting we do in the brain or is it general engagement with a treatment that’s exciting and interesting … that actually leads to these targeted effects? And so I think we clearly see that it’s the targeted algorithms that we have.” Despite the positive results, questions about the product remain. For instance, parents and physicians subjectively perceived about the same amount of improvement in children’s behavior whether they were playing the placebo game or the therapeutic game. And if Akili can get approval, it remains to be seen whether clinicians and insurers will embrace its product. The video game has not been tested head-to-head against ADHD medications or psychotherapy to see if it’s equally effective. © 2017 Scientific American

Keyword: ADHD; Learning & Memory
Link ID: 24396 - Posted: 12.06.2017

Anne Churchland Decisions span a vast range of complexity. There are really simple ones: Do I want an apple or a piece of cake with my lunch? Then there are much more complicated ones: Which car should I buy, or which career should I choose? Neuroscientists like me have identified some of the individual parts of the brain that contribute to making decisions like these. Different areas process sounds, sights or pertinent prior knowledge. But understanding how these individual players work together as a team is still a challenge, not only in understanding decision-making, but for the whole field of neuroscience. Part of the reason is that until now, neuroscience has operated in a traditional science research model: Individual labs work on their own, usually focusing on one or a few brain areas. That makes it challenging for any researcher to interpret data collected by another lab, because we all have slight differences in how we run experiments. Neuroscientists who study decision-making set up all kinds of different games for animals to play, for example, and we collect data on what goes on in the brain when the animal makes a move. When everyone has a different experimental setup and methodology, we can’t determine whether the results from another lab are a clue about something interesting that’s actually going on in the brain or merely a byproduct of equipment differences. © 2010–2017, The Conversation US, Inc.

Keyword: Learning & Memory; Attention
Link ID: 24392 - Posted: 12.05.2017

An analysis of more than 800,000 people has concluded that people who remain single for life are 42 per cent more likely to get dementia than married couples. The study also found that people who have been widowed are 20 per cent more likely to develop the condition, but that divorcees don’t have an elevated risk. Previous research has suggested that married people may have healthier lifestyles, which may help explain the findings. Another hypothesis is that married people are more socially engaged, and that this may protect against developing the condition. The stress of bereavement might be behind the increased risk in those who have been widowed. But marriage isn’t always good for the health. While men are more likely to survive a heart attack if they are married, single women recover better than those who are married. Journal reference: Journal of Neurology, Neurosurgery, and Psychiatry © Copyright New Scientist Ltd.

Keyword: Alzheimers; Learning & Memory
Link ID: 24382 - Posted: 12.01.2017

By JOANNA KLEIN Chances are that’s a shy elk looking back at a bold magpie, in the photograph above. They get along, so to speak, because the elk needs grooming and the magpie is looking for dinner. But they may have never entered into this partnership if it weren’t for their particular personalities, suggests a study published Wednesday in Biology Letters. Let’s start with the elk. In Canada’s western province of Alberta, they’ve been acting strange. Some have quit migrating, opting to hang around towns with humans who protect them from predators like wolves. Others still migrate. As a doctoral student at the University of Alberta, Robert Found, now a wildlife biologist for Parks Canada, discovered over years of observing their personalities that bold elk stayed, while shy elk migrated. But he noticed something else in the process of completing his research: As elk laid down to rest at the end of the day, magpies approached. There appeared to be a pattern: elk of some personality types aggressively rejected magpies. Others didn’t. “Sometimes the magpies will walk around right on the head and the face of the elk,” Dr. Found said. Scientists define animal personality by an individual animal’s behavior. It’s predictable, but also varies from others in a group. Dr. Found created a bold-shy scale for elk, measuring how close they allowed him to get, where elk positioned themselves within the group, which elk fought other elk, which ones won, how long elk spent monitoring for predators and their willingness to approach unfamiliar objects like old tires, skis and a bike. He also noted which elk accepted magpies. To study the magpies, he attracted the birds to 20 experimental sites with peanuts on tree stumps. During more than 20 separate trials with different magpies, he judged each bird’s behavior relative to the other magpies in a trial. Like the elk, he measured flight response, social structure and willingness to approach items they hadn’t previously encountered (a bike decorated with a boa and Christmas ornaments). He also noted who landed on a faux-elk that offered dog food rather than ticks (a previous study showed magpies liked dog food as much as ticks). © 2017 The New York Times Company

Keyword: Learning & Memory; Evolution
Link ID: 24380 - Posted: 11.30.2017

By Julie Hecht A good friend insists: "You don't study dogs, Julie. You study human culture. Dog behavior is a product of the people who love them." And since I'm not one for quick comebacks, I typically just smile and pet her dog. Because I like dogs. And I study them (see what I just did there? Bam). Or maybe she's onto something. Is dog behavior independent from where they live? From the cultural norms they're exposed to? Maybe German Shepherds can tell us a thing or two. In 2009, researchers from Hungary and the USA published a cross-cultural survey where German Shepherd owners from each country weighed in on their dogs. While a number of similarities emerged, so did differences. For example, USA German Shepherds were more likely to be kept indoors and have more types of training experiences. And when it came to behavior, on some measures there was no difference between German Shepherds in each country—all owners reported low activity-impulsivity and low inattention scores—but there were also a few differences: the USA dogs scored higher on confidence and aggressiveness than those in Hungary. Does this mean a German Shepherd here isn't the same as a German Shepherd there? One possible answer is: yes, the dogs are different. If German Shepherd lovers in the USA prefer dogs with higher confidence ratings, this preference "could lead to selective breeding for higher confidence, resulting in a population of German Shepherds in the USA with this trait." We know it’s possible to select for particular parental behavioral traits, and then observe them in offspring. "Genetic isolation, as well as environmental variation, could contribute to differences in pet behavior across cultures," the researchers offer. © 2017 Scientific American

Keyword: Learning & Memory; Evolution
Link ID: 24366 - Posted: 11.27.2017

Laura Sanders In stark contrast to earlier findings, adults do not produce new nerve cells in a brain area important to memory and navigation, scientists conclude after scrutinizing 54 human brains spanning the age spectrum. The finding is preliminary. But if confirmed, it would overturn the widely accepted and potentially powerful idea that in people, the memory-related hippocampus constantly churns out new neurons in adulthood. Adult brains showed no signs of such turnover in that region, researchers reported November 13 at a meeting of the Society for Neuroscience in Washington, D.C. Previous studies in animals have hinted that boosting the birthrate of new neurons, a process called neurogenesis, in the hippocampus might enhance memory or learning abilities, combat depression and even stave off the mental decline that comes with dementia and old age (SN: 9/27/08, p. 5). In rodents, exercise, enriched environments and other tweaks can boost hippocampal neurogenesis — and more excitingly, memory performance. But the new study may temper those ambitions, at least for people. Researchers studied 54 human brain samples that ranged from fetal stages to age 77, acquired either postmortem or during brain surgery. These samples were cut into thin slices and probed with molecular tools that can signal dividing or young cells, both of which are signs that nerve cells are being born. As expected, fetal and infant samples showed evidence of both dividing cells that give rise to new neurons and young neurons themselves in the hippocampus. But with age, these numbers declined. In brain tissue from a 13-year-old, the researchers spotted only a handful of young neurons. And in adults, there were none. |© Society for Science & the Public 2000 - 2017.

Keyword: Neurogenesis
Link ID: 24334 - Posted: 11.16.2017

By Jef Akst | After Nelson Dellis’s grandmother passed away from Alzheimer’s disease in the summer of 2009, he became obsessed with memory. “I had seen her whole decline, so brain health was on my mind,” he says. He found out about annual memory competitions that tested people’s ability to remember large volumes of data—for example, the exact order of 104 playing cards in two decks—and began to learn the strategies so-called “memory athletes” used to pull off these incredible feats. “I found the techniques worked, and with a bit of practice, you can do a lot more than you ever thought you could,” Dellis says. He entered the 2010 USA Memory Championship in New York City and came in third. The next two years in a row, he took first. A mistake in the finals cost him the championship in 2013, but he regained the crown in 2014 and won again in 2015, making him the first and only four-time USA Memory Champion. And all it took was “a bit of practice.” Dellis says there are several strategies memory athletes use, but they’re all based on the same principle: “You want to turn information you’re trying to memorize into something that your brain naturally prefers to absorb”—typically, an image. “Once you have that picture, the next step is to store it somewhere—somewhere in your mind you can safely store it and retrieve it later.” This place is known as a “memory palace,” and it can be any place that’s familiar to you, such as your house. You can then place the images you’ve chosen along a particular path through the memory palace, and “the path, which you know very well, preserves the order.”

Keyword: Learning & Memory
Link ID: 24318 - Posted: 11.11.2017

Using an innovative “NeuroGrid” technology, scientists showed that sleep boosts communication between two brain regions whose connection is critical for the formation of memories. The work, published in Science, was partially funded by the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative, a project of the National Institutes of Health devoted to accelerating the development of new approaches to probing the workings of the brain. “Using new technologies advanced by the BRAIN Initiative, these researchers made a fundamental discovery about how the brain creates and stores new memories,” said Nick Langhals, Ph.D., program director at NIH’s National Institute of Neurological Disorders and Stroke. A brain structure called the hippocampus is widely thought to turn new information into permanent memories while we sleep. Previous work by the new study’s senior author, New York University School of Medicine professor György Buzsáki, M.D., Ph.D., revealed high-frequency bursts of neural firing called ripples in the hippocampus during sleep and suggested they play a role in memory storage. The current study confirmed the presence of ripples in the hippocampus during sleep and found them in certain parts of association neocortex, an area on the brain’s surface involved in processing complex sensory information. “When we first observed this, we thought it was incorrect because it had never been observed before,” said Dion Khodagholy, Ph.D., the study’s co-first author and assistant professor at Columbia University in New York.

Keyword: Learning & Memory
Link ID: 24274 - Posted: 11.01.2017

Jon Hamilton When it comes to brain training, some workouts seem to work better than others. A comparison of the two most common training methods scientists use to improve memory and attention found that one was twice as effective as the other. The more effective method also changed brain activity in a part of the brain involved in high-level thinking. But neither method made anyone smarter, says Kara Blacker, the study's lead author and a researcher at The Henry M. Jackson Foundation for the Advancement of Military Medicine in Bethesda, Md. "Our hypothesis was that training might improve fluid intelligence or IQ," Blacker says. "But that's not what we found." Blacker did the memory research when she was part of a team at Johns Hopkins University and the Kennedy Krieger Institute in Baltimore. The results were reported in the Journal of Cognitive Enhancement. The team compared two approaches to improving working memory, which acts as a kind of mental workspace where we store information temporarily. "If somebody gives you directions, you have to keep that information in mind long enough to actually execute going to that location," Blacker says. "If someone tells you a phone number, you have to be able to remember it." To test different methods for improving working memory, the team had 136 young adults spend a month training their brains for 30 minutes a day, five days a week. Johns Hopkins University YouTube One group did something called a "complex span" test, which involves remembering the location of an item despite distractions. A second group trained with something called the dual n-back test. Each day they would sit at a computer watching flashing squares appear on a grid and listening to a voice reading letters from the alphabet. © 2017 npr

Keyword: Learning & Memory; Attention
Link ID: 24231 - Posted: 10.23.2017

by Sari Harrar, AARP Bulletin, At 99 years old Brenda Milner continues to explore the mind and its relationship to people’'s behavior. You'’re a preeminent neuroscientist, and a professor at Canada's prestigious McGill University. At age 99, what motivates you to keep up your research at the Montreal Neurological Institute and Hospital? I am very curious. Human quirks attract my interest. If you’'re a theoretical person, you can sit and dream up beautiful theories, but my approach is, “What would happen if …”or, “Why is this person doing [that] …”and then, “How can I measure it?” I wouldn't still be working if I didn't find it exciting. AARP Membership: Join or Renew for Just $16 a Year Are you curious in real life, too? Yes. I'm a good "noticer—" of behavior as much as the kind of furniture people have! In the 1950s, you made a revolutionary discovery— that memories are formed in a brain area called the hippocampus, which is now getting lots of attention for its role in memory loss and dementia. Has brain research gotten easier? Nowadays, everyone has functional magnetic resonance imaging. Anybody with access to a medical school can get a good look at the patients' brain while they're alive and young, but it wasn't like that [then]. Psychologists were studying patients who were much older and beginning to show memory impairment. Then they had to wait for their patients to die.

Keyword: Learning & Memory
Link ID: 24196 - Posted: 10.16.2017

By Virginia Morell Dog owners often wonder what—if anything—is going on when their pooches are sleeping. It turns out they may be learning, according to a new study. Researchers in Hungary trained 15 pet dogs to sit and lie down using English phrases instead of the Hungarian they already knew. Afterward, the scientists attached small electrodes to the dogs’ heads to record their brain activity while they slept. Electroencephalograms (EEGs) showed that during 3-hour naps, the dogs’ brains experienced brief, repeated moments of “slow-wave” brain activity, lasting 0.5 to 5 seconds. These bursts—called sleep spindles because they look like a train of fast, rhythmic waves on EEG recordings—occur during non-REM sleep and are known to support memory, learning, general intelligence, and healthy aging in humans and rats. But this is the first time they’ve been studied in detail in dogs. Like those of humans and rats, the dogs’ sleep spindles occur in short cycles in the 9-hertz to 16-hertz range; in humans and rats, these cycles are associated with memory consolidation. The scientists also discovered that the number of spindle sessions per minute correlated with how well the dogs learned their new, foreign vocabulary, the researchers report this week in Scientific Reports. And—just like in humans—females had more spindle sessions per minute than males and performed better during testing. About 30% of the females learned the new words, compared to about 10% of the males. That suggests, the researchers say, that dogs can serve as models to better understand the function of our own sleep spindles. © 2017 American Association for the Advancement of Science

Keyword: Sleep; Learning & Memory
Link ID: 24191 - Posted: 10.14.2017

Laura Sanders The brain’s mapmakers don’t get a break, even for sleep. Grid cells, specialized nerve cells that help keep people and other animals oriented, stay on the clock 24/7, two preliminary studies on rats suggest. Results from the studies, both posted October 5 at bioRxiv.org, highlight the stability of the brain’s ‘inner GPS’ system. Nestled in a part of the brain called the medial entorhinal cortex, grid cells fire off regularly spaced signals as a rat moves through the world, marking a rat’s various locations. Individual grid cells work together to create a mental map of the environment. But scientists didn’t know what happens to this map when an animal no longer needs it, such as during sleep. Grid cells, it turns out, maintain their mapmaking relationships even in sleeping rats, report two teams of researchers, one from the University of Texas at Austin and one from the Norwegian University of Science and Technology in Trondheim. (The Norway group includes the researchers who won a Nobel Prize in 2014 for discovering grid cells (SN Online: 10/6/14).) By eavesdropping on pairs of grid cells, researchers found that the cells maintain similar relationships to each other during sleep as they do during active exploration. For instance, two grid cells that fired off signals nearly in tandem while the rat was awake kept that same pattern during sleep, a sign that the map is intact. The results provide insights into how grid cells work together to create durable mental maps. © Society for Science & the Public 2000 - 2017.

Keyword: Learning & Memory
Link ID: 24186 - Posted: 10.13.2017

Children with attention deficit hyperactivity disorder may fidget, tap and swivel around in a chair much more than normally developing children because it helps them to learn complex material, psychologists have found. ADHD is often perceived as a behavioural problem because it can result in symptoms such as inattention, impulsivity, and hyperactivity that can affect social interaction and learning. Scientists increasingly recognize ADHD as a brain disorder that affects about five per cent of the school-age population. Now brain tests show children with ADHD tend to learn less when sitting still compared to when they're moving. It is not for lack of motivation, says Prof. Mark Rapport, a child psychopathology researcher who focuses on ADHD at the University of Central Florida in Orlando. Rapport and his colleagues set out to test an observation made by many parents — that children with ADHD can pay attention if they are doing an activity they enjoy. They put 32 boys aged eight to 12 with ADHD and 30 of their peers who are not affected by the disorder through a battery of memory and other tests. Participants watched two videos on separate days: an instructional math lesson without performing the calculations, and a scene from Star Wars Episode 1 — The Phantom Menace. During the Star Wars movie, the boys with ADHD did not squirm more than other children, but when asked to concentrate on the math lesson, there was a difference between the two groups. "All children and all people in general, moved more when they were engaged in a working memory task. Kids with ADHD move about twice as much under the same conditions," Rapport said. ©2017 CBC/Radio-Canada.

Keyword: ADHD; Learning & Memory
Link ID: 24164 - Posted: 10.09.2017

By Giorgia Guglielmi This mantis shrimp (Gonodactylus smithii) might have a much more elaborate brain than previously thought. That’s the conclusion of the first study to peer into the head of more than 200 crustaceans, including crabs, shrimp, and lobsters. Researchers discovered that the brain of mantis shrimp contains memory and learning centers, called mushroom bodies, which so far have been seen only in insects. The team also found similar structures in close relatives of these sea creatures: cleaner shrimp, pistol shrimp, and hermit crabs. This may not be a coincidence, the researchers say, because mantis shrimp and their brethren are the only crustaceans that hunt over long distances and might have to remember where to get food. But the finding, reported in eLife, is likely to stir debate: Scientists agree that mushroom bodies evolved after the insect lineage split off from the crustacean lineage about 480 million years ago; finding these learning centers in mantis shrimp means that either mushroom bodies are much more ancient than scientists realized and were lost in all crustaceans but mantis shrimp, or that these structures are similar to their counterparts in insects but have evolved independently. © 2017 American Association for the Advancement of Science.

Keyword: Learning & Memory; Evolution
Link ID: 24158 - Posted: 10.07.2017

By Clare Wilson OUR braininess may have evolved thanks to gene changes that made our brain cells less sticky. The cortex is the thin, highly folded outer layer of our brains and it is home to some of our most sophisticated mental abilities, such as planning, language and complex thoughts. Around three millimetres thick, this layer is folded into an intricate pattern of ridges and valleys, which allows the cortex to be large, but still fit into a relatively small space. Many larger mammals, such as primates, dolphins and horses, have various patterns of folds in their cortex, but folds are rarer in smaller animals like mice. So far, we have only identified a few genetic mutations that contributed to the evolution of the human brain, including ones that boosted the number of cells in the cortex. One theory about how the cortex came to be folded is that it buckled as the layer of cells expanded. Daniel del Toro at the Max Planck Institute of Neurobiology in Munich, Germany, and colleagues wondered if some of the genetic changes in our brain’s evolution might have been about more than just an increasing number of cells. They investigated the genes for two molecules – FLRT1 and FLRT3 – which make developing brain cells stick to each other more. Human brain cells produce only a small amount of these compounds, while mice brain cells make lots. Del Toro’s team created mice embryos that lacked functioning FLRT1 and FLRT3 genes, which meant their cortex cells were only loosely attached to each other, like those of humans. © Copyright New Scientist Ltd.

Keyword: Development of the Brain; Learning & Memory
Link ID: 24153 - Posted: 10.05.2017

By Jessica Hamzelou AT LAST, we’ve seen how the brain memories when we sleep. By scanning slumbering people, researchers have watched how the “trace” of a memory moves from one region of the brain to another. “The initial memory trace kind of disappears, and at the same time, another emerges,” says Shahab Vahdat at Stanford University in California. It is the first time memories have been observed being filed away in humans during sleep, he says. Vahdat and his colleagues did this by finding people who were able to fall asleep in the confined, noisy space of an fMRI scanner, which is no easy undertaking. “We screened more than 50 people in a mock scanner, and only 13 made it through to the study,” says Vahdat. The team then taught this group of volunteers to press a set of keys in a specific sequence – in the same way that a pianist might learn to play a tune. It took each person between about 10 and 20 minutes to master a sequence involving five presses. “They had to learn to play it as quickly and as accurately as possible,” says Vahdat. Once they had learned the sequence, each volunteer put on a cap of EEG electrodes to monitor the electrical activity of their brain, and entered an fMRI scanner – which detects which regions of the brain are active. The team saw a specific pattern of brain activity while the volunteers performed the key-pressing task. Once they had stopped, this pattern kept replaying, as if each person was subconsciously revising what they had learned. © Copyright New Scientist Ltd.

Keyword: Sleep; Learning & Memory
Link ID: 24151 - Posted: 10.05.2017

By Claudia Wallis, A funny thing happened in the Dutch city of Maastricht in the fall of 2011. A policy went into effect banning the sale of marijuana at the city’s 13 legal cannabis shops to visitors from most other countries. The goal was to discourage disruptive drug tourism in a city close to several international borders. The policy had its intended effect, but also a remarkable unintended one: foreign students attending Maastricht University starting getting better grades. According to an analysis published earlier this year in Review of Economic Studies, students who had been passing their courses at a rate of 73.9% when they could legally buy weed were now passing at a rate of 77.9% — a sizeable jump. The effect, which was based on data from 336 undergraduates in more than 4,000 courses, was most dramatic for weaker students, women, and in classes that required more math. Some of this falls in line with past research: marijuana use has been linked to inferior academic achievement (and vice versa), so it makes sense that poorer students might benefit most from a ban, and the drug is known to have immediate effects on cognitive performance, including in math. But what’s really unusual about the study, notes one of its authors, economist Ulf Zoelitz of the Briq Institute on Behavior and Inequality, is that rather than merely correlating academic performance with cannabis use, as much past research has done, “we could cleanly identify the causal impact of a drug policy.” Zoelitz co-authored the study with Olivier Marie of Erasmus University Rotterdam. © 2017 KQED Inc.

Keyword: Learning & Memory; Drug Abuse
Link ID: 24139 - Posted: 10.03.2017

By Matthew Hutson Studying the human mind is tough. You can ask people how they think, but they often don’t know. You can scan their brains, but the tools are blunt. You can damage their brains and watch what happens, but they don’t take kindly to that. So even a task as supposedly simple as the first step in reading—recognizing letters on a page—keeps scientists guessing. Now, psychologists are using artificial intelligence (AI) to probe how our minds actually work. Marco Zorzi, a psychologist at the University of Padua in Italy, used artificial neural networks to show how the brain might “hijack” existing connections in the visual cortex to recognize the letters of the alphabet, he and colleagues reported last month in Nature Human Behaviour. Zorzi spoke with Science about the study and about his other work. This interview has been edited for brevity and clarity. Q: What did you learn in your study of letter perception? A: We first trained the model on patches of natural images, of trees and mountains, and then this knowledge becomes a vocabulary of basic visual features the network uses to learn about letter shapes. This idea of “neural recycling” has been around for some time, but as far as I know this is the first demonstration where you actually gained in performance: We saw better letter recognition in a model that trained on natural images than one that didn’t. Recycling makes learning letters much faster compared to the same network without recycling. It gives the network a head start. © 2017 American Association for the Advancement of Science.

Keyword: Learning & Memory; Robotics
Link ID: 24130 - Posted: 09.30.2017