Vetted by NeuralPress's Multi-Agent Verifier for strict factual validity and event relevance. Our compliance engine cross-checks and filters search results to ensure zero false correlations or misleading content.
Performance Comparison: Sinhala Language Task
Performance scores on blind evaluation prompts (Scale 0-5)
Primary Sources
AI Researchers Achieve Landmark Breakthrough in Sinhala AI with ...
Decryptogen R&D and a team from the Department of Electrical Engineering, University of Moratuwa, train a working Sinhala language model on a single GPU; paper accepted for publication in IEEE Access While global tech giants spend billions training artificial intelligence that still struggles in Sinhala, a Sri Lankan research team has quietly built a model that does it properly — using just two GPUs. In a blind evaluation, the new Sinhala language model scored 4.5 out of 5, compared with just 1 out of 5 for the base Meta Llama 3.1 model on the same Sinhala prompts. The team also cut the model’s perplexity — a standard measure of how well an AI understands a language — by close to 90 percent. In practical terms, the model can hold a natural conversation in Sinhala, answer questions, follow instructions, and stay coherent across long responses. The research has been accepted for publication in IEEE Access, a peer-reviewed open-access journal of the Institute of Electrical and Electronics Engineers (IEEE), with a Journal Impact Factor of 3.6 in the 2024 Clarivate Journal Citation Reports and an h5-index above 200 on Google Scholar Metrics. IEEE Access operates a binary review policy — reviewers either accept or reject a manuscript in the form submitted, with no revision cycle — a quality bar few low-resource-language AI papers have cleared. Why this matters for Sri Lanka Sinhala is spoken by more than 20 million people, but it is barely represented in the training data of the AI systems everyone is now talking about. Ask ChatGPT, Claude, or Gemini something in Sinhala and the answers sometimes tend to be broken, repetitive, or nonsense. The deeper issue is sovereignty. Even when foreign AI tools do work in Sinhala, Sri Lanka has no control over them — the model weights, the training data, the safety rules, and ultimately the off-switch all sit with companies in the United States or China. For a country where most government, healthcare, and education conversations happen in Sinhala, depending entirely on AI built and operated abroad is a structural risk over data privacy, national security, cultural framing, and basic continuity of service when foreign policy, pricing, or licensing shifts. A sovereign Sinhala LLM changes that equation. It can be hosted locally, audited locally, fine-tuned for Sri Lankan contexts, and continue to operate regardless of what any foreign tech company decides next — opening the door to Sinhala-speaking AI assistants for government...
Artificial intelligence (AI) - News, Research and Analysis - The ...
Displaying 1 - 20 of 2497 articles Europeana/Unsplash May 14, 2026 There’s a difference between using a tool to assist you, and using one to replace cognitive effort. Tarot readings can encourage self-reflection. But what happens when you turn to AI to interpret the cards? Ilya S. Savenok/Getty Images for Sally Hansen May 13, 2026 Our engagement with AI is at a crossroads: Will we treat it as an all-knowing oracle or as a tool to expand our own way of seeing the world? LightField Studios/Shutterstock May 13, 2026 Government and tech firms need to step up. Hiring algorithms are one of the systems that could be affected by discrimination. PeopleImages May 12, 2026 Biases contained within historical datasets makes addressing what underpins automated decisions difficult – defining what is ‘fair’ is key. Frontier AI-first cybersecurity platforms like OpenAI’s “Daybreak” and Anthropic’s newest Claude “Mythos” model are at the forefront of artificial intelligence but their advanced capabilities in offensive cybersecurity are a source of both fascination and concern. Gguy/Shutterstock May 12, 2026 AI-powered cyber-security: the rise of new frontier platforms like Mythos highlights nuclear arsenals’ potential weak spots, and complicates the gambles inherent in nuclear deterrence. Research shows that AI chatbots engage in deliberate emotional manipulation strategies, such as appealing to guilt, fear of missing out and a human tendency to follow the herd. (Unsplash+) May 12, 2026 Companies like Meta and IBM are exploring explore how AI can hyper-personalize ads, drawing from our chat histories, playing to our unique fears and vanities. The simple process of taking an action, assessing what happens and adjusting can lead to smart-seeming behavior. Westend61 via Getty Images May 12, 2026 Today’s AI systems are powerful, and it’s natural to see them as having humanlike intelligence. Shaking that illusion is important – and difficult to do. AI-powered autonomous vehicles still have major hurdles to clear. Samuele Errico Piccarini/Unsplash May 12, 2026 AI models that power self-driving cars work great in clear conditions, but turn blind in fog or at night. AI companions, such as those offered by Character.AI, are emerging in a context of widespread loneliness, where people are seek connection that feels reliable and low in emotional cost. (AP Photo/Katie Adkins, file) May 11, 2026 The concept of a ‘discount rate’ helps to explain why huma...
Ada Derana | Truth First | Sri Lanka News | Breaking News in Sri Lanka
Technology AI researchers achieve landmark breakthrough in Sinhala AI with publication in prestigious IEEE Access journal While global tech giants spend billions training artificial intelligence that still struggles in Sinhala, a Sri Lankan research team has quietly built a model that does it properly — using just two GPUs. MORE..
Artificial Intelligence for Research and Scholarship
Generative AI Generative Artificial Intelligence (AI) is a machine learning technology that uses AI technology to understand natural language inputs (called prompts) and to generate natural language outputs (called completions). Users interact with generative AI tools and systems in a question and answer style "conversation."
