Robotics
What’s new Meet SENTINEL, an end-to-end model that lets humanoid robots turn plain-language instructions into whole-body motion - no hand-crafted pipelines or teleop in the loop. How it works * Trains on a large dataset built by tracking human motions in simulation and pairing them with text descriptions. * Directly maps
AI
What if your watch could help estimate your daily calories—not just count steps? Researchers introduce the Nutrition Photoplethysmography Language Model (NPLM), which blends heart-signal data from consumer wearables (PPG) with meal descriptions so AI can reason about both physiology and food. Trained on 19,340 people and 1.1
AI
Ever notice how some AIs agree with a confident (but wrong) authority? That’s sycophancy. PARROT is a new benchmark that tests how much language models bend under authority and persuasion. * How it works: The same question is asked two ways—neutral vs. with an authoritative (but wrong) cue—scored
NLP
Smarter news sorting Classifying news articles is tricky: simple stats like term frequency catch key words, while deep models capture context. AGFF (Attention‑Guided Feature Fusion) blends both—and uses attention to decide, per article, which signal to trust more. * Fuses statistical features (e.g., TF‑IDF) with semantic representations
AI
AI that understands your preferences PersonaAgent with GraphRAG is a new way to build language model agents that adapt to you. The agent learns a compact "persona" from your past choices and behaviors, stored in a knowledge graph. It also finds communities of related information across documents to
AI
Teach Robots to Learn Like People Building versatile embodied AI (agents that see, read, and act) is hard because real-world data is scarce and training is expensive. We introduce Deliberate Practice Policy Optimization (DPPO), a metacognitive "metaloop" that alternates between: * Supervised fine-tuning to expand general competence * Reinforcement learning
AI
NutriScreener is a new AI method to help screen for child malnutrition quickly and at scale—using ordinary photos captured from multiple poses. How it works: it combines CLIP-based visual features, a retrieval “knowledge base” that surfaces context from similar cases, and a multi-pose graph attention network that reasons across
AI
AI images that "think" as they form Most image generators plan before they draw or fix mistakes after. Thinking-while-Generating (TwiG) lets models do something new: interleave short bursts of textual reasoning during generation—like thinking out loud while painting. As pixels appear, the model explains what to add
AI
Can we quickly tell if an image is AI-made — and even which generator produced it — without retraining huge models every month? This study says yes. The authors propose a two-stage system: * Learn a visual "fingerprint" space. A vision model trained with supervised contrastive learning separates real vs. synthetic
AI
Building capable embodied AI is hard: real-world data is scarce and training is expensive. This paper introduces Deliberate Practice Policy Optimization (DPPO), a coach-like training loop that helps vision-language agents learn more from less. How it works: the system alternates between learning from examples (to expand skills) and trial-and-error reinforcement
AI
Wearables are reshaping healthcare, but most AI models track one physical condition at a time and overlook depression—despite frequent overlap. This study reframes multi-disease assessment as a multi-task learning problem, letting one model learn several conditions jointly and use their relationships to improve accuracy. The catch is “double heterogeneity”
AI
Wearables and AI are reshaping healthcare, but most tools track only physical diseases. That misses depression, which often coexists with chronic conditions. This study introduces one model that assesses multiple chronic diseases and depression together, using data from everyday sensors. It treats each disease as its own task, so the