In Partially Observable Markov Decision Processes, integrating an agent's history into memory poses a significant challenge for decision-making. Traditional imitation learning, relying on observation-action pairs for expert demonstrations, fails to capture the expert's memory mechanisms used in decision-making. To capture memory processes as demonstrations, we introduce the concept of memory dependency pairs $(p, q)$ indicating that events at time $p$ are recalled for decision-making at time $q$. We introduce AttentionTuner to leverage memory dependency pairs in Transformers and find significant improvements across several tasks compared to standard Transformers when evaluated on Memory Gym and the Long-term Memory Benchmark. Code is available at https://github.com/WilliamYue37/AttentionTuner.

Fox News medical contributor Dr. Janette Nesheiwat joins'Fox News Live' to discuss the results of an Alzheimer's drug trial by Eli Lilly. Alzheimer's is a disease that heavily impacts memory function, typically in older people. Since memory loss commonly comes with aging, it can be hard to detect if a symptom is just one that comes with old age, or a sign of Alzheimer's disease. The main symptom of Alzheimer's disease is memory loss. More than that, it's memory loss that happens frequently and gets worse over time.

Large Transformer-based Pretrained Language Models (PLMs) dominate almost all Natural Language Processing (NLP) tasks. Nevertheless, they still make mistakes from time to time. For a model deployed in an industrial environment, fixing these mistakes quickly and robustly is vital to improve user experiences. Previous works formalize such problems as Model Editing (ME) and mostly focus on fixing one mistake. However, the one-mistake-fixing scenario is not an accurate abstraction of the real-world challenge. In the deployment of AI services, there are ever-emerging mistakes, and the same mistake may recur if not corrected in time. Thus a preferable solution is to rectify the mistakes as soon as they appear nonstop. Therefore, we extend the existing ME into Sequential Model Editing (SME) to help develop more practical editing methods. Our study shows that most current ME methods could yield unsatisfying results in this scenario. We then introduce Transformer-Patcher, a novel model editor that can shift the behavior of transformer-based models by simply adding and training a few neurons in the last Feed-Forward Network layer. Experimental results on both classification and generation tasks show that Transformer-Patcher can successively correct up to thousands of errors (Reliability) and generalize to their equivalent inputs (Generality) while retaining the model's accuracy on irrelevant inputs (Locality). Our method outperforms previous fine-tuning and HyperNetwork-based methods and achieves state-of-the-art performance for Sequential Model Editing (SME). The code is available at https://github.com/ZeroYuHuang/Transformer-Patcher.

Like our brains, a new computer program created by Parham Aarabi of the University of Toronto can store and retrieve information strategically. An experimental tool that uses the novel algorithm to aid those with memory loss has also been developed by the associate professor in the Faculty of Applied Science & Engineering's Edward S. Rogers Sr. department of electrical and computer engineering. In the minds of most people, AI is more robotic than humans, according to Aarabi, whose approach is examined in a paper presented at the IEEE Engineering in Medicine and Biology Society Conference in Glasgow. Aarabi believes it should change. Computers have traditionally needed explicit instructions from their users on what data to save.

A new computer algorithm developed by the University of Toronto's Parham Aarabi can store and recall information strategically – just like our brains. The associate professor in the Edward S. Rogers Sr. department of electrical and computer engineering, in the Faculty of Applied Science & Engineering, has also created an experimental tool that leverages the new algorithm to help people with memory loss. "Most people think of AI as more robot than human," says Aarabi, whose framework is explored in a paper being presented this week at the IEEE Engineering in Medicine and Biology Society Conference in Glasgow. "I think that needs to change." In the past, computers have relied on their users to tell them exactly what information to store.

Instead of our memories decaying with time, forgetting is actually an active form of learning that helps our brain to access more important information. This is the conclusion of experts from Trinity College Dublin and the University of Toronto, who said that'lost' memories are not really gone, just made inaccessible. Memories, they explained, are stored permanently in sets of neurons, with our brains deciding which ones we keep access to and which irrelevant ones are locked away. These choices, they said, are based on environmental feedback, theoretically allowing us flexibility in the face of change and better decision-making as a result. If correct, the findings could lead to new ways to understand and treat memory loss associated with disease -- such as is seen, for example, in patients with Alzheimer's.

Artificial intelligence and the cell's self-cleansing system are the keys behind the novel medication. The treatment may strengthen other organs as well. One in six Norwegians over 80 is affected by Alzheimer's disease. Numbers are even higher worldwide, and there is still no cure available. Researchers at the faculty have developed an artificial intelligence (AI) method to help them identify potential new medicines for Alzheimer's.

Dementia is the deterioration of cognitive functioning thinking, remembering, problem-solving and reasoning which can interfere with daily life. Though dementia is more common in older adults, it is not a part of normal aging. It can also affect younger people. Are you at risk of developing dementia? Artificial intelligence can predict that, concluded a study published in JAMA Network Open.

Mammals including humans are more likely to suffer from short-term memory loss during the winter, a new study suggests. Experts at the University of Bradford looked at how rats performed in memory tests when exposed to long and short'photoperiods', or periods of exposure to light. They found a'significant' link between poor memory and short day length, akin to what humans experience during the winter season. It's possible that the results could be applicable to humans, the team say, suggesting we're more prone to forgetfulness during long winters. Short-term memory loss refers to forgetting things that occurred recently, such as conversations or events.

Jacqui Quibbell has suffered from'crippling periods of depression and suicidal thoughts' for all her adult life. In 2003, her doctors suggested Jacqui underwent electro-convulsive therapy (ECT). This involves attaching electrodes to the patient's head and, under general anaesthetic, passing electric shocks through their brain -- which is said to'rewire' it. 'I didn't know much about ECT, I didn't have Google then,' says Jacqui, 57. 'I started suffering memory loss during the treatment and by the time it finished, my short-term memory had disappeared completely and has never come back.