LLMs are great at step-by-step "chain-of-thought" logic, but real breakthroughs need creative reasoning. This paper introduces Universe of Thoughts (UoT): a framework, inspired by cognitive science, that teaches AI to explore many idea paths—not just the obvious ones. * Combinational: mix existing ideas in novel ways. * Exploratory: push
What if computers could not only see, but also smell? Researchers introduce New York Smells, a large, real-world dataset pairing photos with olfactory sensor readings. The set spans 7,000 smell-image pairs from 3,500 different objects across indoor and outdoor environments—about 70x more objects than prior smell datasets.
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
How do we know an AI is truly helpful, honest, safe—and follows instructions? Today, we usually read its answers and score them, which can be slow, costly, and subjective. This paper finds a simple clue: when compared to a trusted reference, models that write well also judge well. The
AI
LLMs can write tidy code that passes unit tests—but can they build money-making agents that plan, bid, and deliver under pressure? This study pitted 40 LLM-coded agents (prompted with methods including “vibe coding,” i.e., high-level guidance) against 17 human-coded agents from grad CS students in a real-world-style challenge:
AI
Can AI write like Dickens with a single word? What if one token could cue an AI to channel Austen, Twain, or Melville? This study fine-tunes large language models so a minimal, single-token prompt triggers 19th-century novelist styles. * How they checked it: An AI detector trained on authentic sentences judged
TL;DR AI can now draft UIs and slides, but design taste is personal. DesignPref is a new dataset that captures individual preferences from professional designers—so models can learn your style, not the crowd’s. * What it is: 12k pairwise comparisons of AI-generated UIs, rated by 20 pros with
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
AI
Accuracy isn’t the same as reliability. Meet PaTAS—a “trust meter” that runs in parallel with neural networks to quantify how much you should trust each prediction, using Subjective Logic. * How it works: PaTAS adds Trust Nodes and Trust Functions that propagate trust about inputs, parameters, and activations. A
AI
TL;DR An open-source platform helps creators check if their content was used to train large language models, without heavy compute or opaque workflows. What it does * Verifies if a specific work appears in LLM training datasets. * Improves similarity matching and dataset validation. * Cuts compute by 10-30% via efficient API
What if you could change how a video moves—without changing how it looks? MotionV2V lets editors "edit motion, not pixels." The team defines a motion edit as a small tweak to the sparse paths (trajectories) objects follow—like a hand wave, a camera pan, or a dancer&
Can AI grasp figurative language rooted in culture? BengaliFig puts it to the test for Bengali, a widely spoken yet low-resourced language. This compact, carefully annotated challenge set gathers 435 riddles from Bengali oral and literary traditions and turns them into multiple-choice questions via a constraint-aware, AI-assisted pipeline. * 5 annotation