AI
Can we reliably spot AI-written text? This study says yes—when we evaluate detectors the right way. Researchers built a unified, fair benchmark for AI-text detection by combining HC3 and DAIGT v2 and using a topic-based split to prevent “memorizing” subjects. That makes models prove they generalize to new domains.
AI
AI assistants that click, type, and navigate apps need practice worlds — but real, trainable web environments are scarce. InfiniteWeb changes that by automatically generating realistic, multi-page websites at scale so GUI agents can learn real tasks, not toy demos. How it works * Unified specs: A single blueprint ties pages, flows,
AI
LLMs can learn to reason—without task verifiers Many real-world problems don’t have automatic checkers to grade answers, even though we have lots of expert solutions. RARO (Relativistic Adversarial Reasoning Optimization) shows how to train reasoning skills from those examples alone. How it works: * A policy (the model) tries
AI
Large models are powerful, but solving deep, multi-step problems is pricey. ToolOrchestra trains a small “orchestrator” that decides which tools to use, when, and how—like a smart project manager for AI. It uses reinforcement learning with rewards for outcomes, efficiency, and user preferences. The result: an 8B Orchestrator that
LLM
What’s new Adding the right metadata can make LLM pretraining faster and more effective—and it’s not just about URLs. * Fine‑grained signals work: Prepending detailed quality indicators helps models learn quicker. * Append-and-predict: Appending metadata and training the model to predict it as an auxiliary task boosts efficiency.
AI
AI on our phones still makes awkward mistakes—and fixing them usually means heavy retraining in the cloud. This research shows a lightweight way to correct errors on-device, in seconds. What’s new: * One-shot fixes: Show the app a single example, and it updates its prototype for that class—no
AI
How do we know if an AI model is truly aligned with human preferences—helpful, honest, safe, and instruction-following? This paper explores a surprisingly effective shortcut: judge the judges. Instead of grading a model’s open-ended answers (which needs lots of human effort or very strong AI judges), the authors
AI
Can AI think creatively? Large language models are great at step-by-step logic, but real-world breakthroughs need more than routine problem-solving. This paper proposes a creative reasoning framework, rooted in cognitive science, and turns it into practice with "Universe of Thoughts" (UoT): methods that help LLMs search wider and
AI
AI assistants are starting to browse the web for us—but webpages can hide sneaky instructions that make them misbehave. This paper maps that threat, called prompt injection, and shows how to defend against it in real-world browsing. What the researchers did * Built a realistic benchmark of attacks embedded in
AI
Computers can see and hear, but smelling the world has been out of reach. New research introduces New York Smells: a large, real-world dataset that pairs odors with images. * 7,000 smell-image pairs from 3,500 objects, indoors and outdoors — about 70x more objects than past datasets. * Benchmark tasks: match
QuantumComputing
What if we could let evolution design quantum circuits that compress data? This paper explores exactly that with quantum autoencoders—models that shrink high‑dimensional quantum or classical data while keeping the important bits. The authors introduce a neural architecture search framework powered by a genetic algorithm. Instead of hand‑
AI
Meet MAESTRO: an AI coach for teamwork Training teams of AI agents is hard: you must design the right "score" (rewards) and the right practice plan (curriculum). Get either wrong and learning stalls. MAESTRO uses a large language model (LLM) as an offline training architect—so it guides