Transformer-based large language models (LLMs) have achieved strong performance across many natural language processing tasks. Nonetheless, their quadratic computational and memory requirements, particularly in self-attention layers, pose challenges for efficient inference on long contexts and for deployment in resource-limited environments. We present TPTT (Transforming Pretrained Transformers into Titans), a framework designed to augment pretrained Transformers with linearized attention (LiZA) and internal memory gating via Memory as Gate (MaG), applied without full retraining. TPTT supports parameter-efficient fine-tuning (LoRA) and integrates with standard toolkits such as Hugging Face Transformers. We evaluated TPTT on several pretrained models, including Llama-1B, OlMoE-1B-7B, Qwen2.5-1.5B, Gemma3-270m, OpenELM-1.3B, and Mistral-7B, in order to assess applicability across architectures of different scales. Experiments on models with approximately 1 billion parameters, evaluated primarily on the MMLU benchmark, suggest potential improvements in both efficiency and accuracy compared to baseline models. For example, Titans-Llama-1B exhibited up to a 20\% relative increase in Exact Match scores in one-shot evaluation. An additional finding is that it is possible to convert a quadratic-attention model into a purely linear-attention model using the DeltaProduct mechanism. All training runs were carried out with modest computational resources. These preliminary findings indicate that TPTT may help adapt pretrained LLMs for long-context tasks with limited overhead. Further studies on larger models and a broader set of benchmarks will be necessary to evaluate the generality and robustness of the framework. Code is available at https://github.com/fabienfrfr/tptt . Python package at https://pypi.org/project/tptt/ .

We study the problem of reconstructing tabular data from aggregate statistics, in which the attacker aims to identify interesting claims about the sensitive data that can be verified with 100% certainty given the aggregates. Successful attempts in prior work have conducted studies in settings where the set of published statistics is rich enough that entire datasets can be reconstructed with certainty. In our work, we instead focus on the regime where many possible datasets match the published statistics, making it impossible to reconstruct the entire private dataset perfectly (i.e., when approaches in prior work fail). We propose the problem of partial data reconstruction, in which the goal of the adversary is to instead output a $\textit{subset}$ of rows and/or columns that are $\textit{guaranteed to be correct}$. We introduce a novel integer programming approach that first $\textbf{generates}$ a set of claims and then $\textbf{verifies}$ whether each claim holds for all possible datasets consistent with the published aggregates. We evaluate our approach on the housing-level microdata from the U.S. Decennial Census release, demonstrating that privacy violations can still persist even when information published about such data is relatively sparse.

Entity linking models have achieved significant success via utilizing pretrained language models to capture semantic features. However, the NIL prediction problem, which aims to identify mentions without a corresponding entity in the knowledge base, has received insufficient attention. We categorize mentions linking to NIL into Missing Entity and Non-Entity Phrase, and propose an entity linking dataset NEL that focuses on the NIL prediction problem. NEL takes ambiguous entities as seeds, collects relevant mention context in the Wikipedia corpus, and ensures the presence of mentions linking to NIL by human annotation and entity masking. We conduct a series of experiments with the widely used bi-encoder and cross-encoder entity linking models, results show that both types of NIL mentions in training data have a significant influence on the accuracy of NIL prediction. Our code and dataset can be accessed at https://github.com/solitaryzero/NIL_EL

The Suite Spot is a fireside chat about all topics IT and OT. We will attempt to bring clarity to the business value of traditionally tech topics. We remove the fog of acronym war and deliver to you the value you need to make these complex technologies work for your business.Powered by RedCircleAs the pandemic continues to have a global impact on business, some are thriving while others are struggling. To think about the "business" side of the pandemic, Brian Householder, President of Digital Infrastructure at Hitachi Vantara, shared what he has learned within his own organization and from other leaders. "I've been talking about the challenges and fielding questions from others. What's different about this situation is that it's a shared experience. The biggest questions about business during and after the pandemic is how this will change how people work, behave, consume, and do business. It's a huge paradigm shift for many, supported by agility.Householder said, "The creativity of the team is amazing.

So why has the process of taking out an insurance policy – and making a claim – become so impersonal? The average home now houses contents worth £35,000, according to the Association of British Insurers – nearly £1 trillion in total. And that doesn't include the value of property itself. With the cost of fire, theft or water damage so high, it is no wonder householders choose to take control of the risk of damage, by taking out home insurance. But, sometimes, consumers feel like the partnership with their insurer is unbalanced and that the supplier holds all the cards.

The Fourth Industrial Revolution, or the marriage of physical and digital technologies, promises to upend how all of us work, from interns to top executives. And yet, despite the clear impact Industry 4.0 will have on workforces in every industry and geography, many executives do not express urgency when it comes to tackling the challenge of the future of the workforce, according to a new report from Forbes Insights and Deloitte, "The Fourth Industrial Revolution Is Here--Are You Ready?" Inevitably, technology is resulting in some human skills becoming obsolete. Based on a survey of more than 1,600 executives worldwide, the report notes that talent/HR is at the very bottom of the list of issues executives focus on, and only 22% of respondents believe that the uncertain impact of Industry 4.0 on the workforce will be one of the top issues affecting their organizations. At the same time, however, only a quarter of executives surveyed express high confidence that they have the right workforce composition and skill sets needed for the future. Despite these contradictions in outlook, 86% of executives believe they are doing everything they can to create a workforce for the Fourth Industrial Revolution.

LG has begun this year's Consumer Electronics Show in Las Vegas the way the show is expected to continue all week: the Korean electronics giant talked about almost nothing other than artificial intelligence the entire time. At the first major press conference at the world's largest annual gadget festival, LG showed how artificial intelligence would be used in its washing machines to adjust cycles depending on the weather, the air quality and whether the householder has exercise time blocked out in his or her calendar. It showed how AI would appear in its air conditioners, learning the behaviour of people in the house so it knows when to turn itself off and on, just like a Nest smart thermostat already does. It showed how AI in its robot vacuums would learn to distinguish householders from furniture, and how AI in its TVs would be used to improve the image quality in its OLED TVs. And, as widely expected, it announced a deal with Google, in which the search giant's artificial intelligence platform, Google Assistant, would be embedded in LG TVs later this year, allowing users to control all sorts of appliances in their house just by talking to Google Assistant on the TV.

Light globes that change colour with the tap of an app, coffee machines you can talk to, and ovens that know exactly how long to cook your food: our homes are getting smart. These devices, just a few examples of what is known as "the internet of things" (or IOT), have been called the "next great disruptor" and "the second digital revolution". One of the great hopes of this revolution is that it will help households save energy. Sensors can turn off lights and appliances when not in use, or turn the heating down when people go to bed. Smartphone apps can provide households with more insight into the energy use of their appliances.

Restructuring electricity grids to meet the increased demand caused by the electrification of transport and heating, while making greater use of intermittent renewable energy sources, represents one of the greatest engineering challenges of our day. This modern electricity grid, in which both electricity and information flow in two directions between large numbers of widely distributed suppliers and generators — commonly termed the ‘smart grid’ — represents a radical reengineering of infrastructure which has changed little over the last hundred years. However, the autonomous behaviour expected of the smart grid, its distributed nature, and the existence of multiple stakeholders each with their own incentives and interests, challenges existing engineering approaches. In this challenge paper, we describe why we believe that artificial intelligence, and particularly, the fields of autonomous agents and multi-agent systems are essential for delivering the smart grid as it is envisioned. We present some recent work in this area and describe many of the challenges that still remain.