Papers200

The paper studies whether AI models can hide or reshape their step-by-step thoughts (chains of thought) on command.

This paper introduces MM-Lifelong, a 181-hour, multi-scale video dataset designed to test AI on true long-term (lifelong) understanding across days to months.

Reasoning models often talk too much, and those extra words can actually make them more wrong.

KARL is a smart search helper that learns to look up information step by step and explain answers using the facts it finds.

This paper says we should test AI the way real life works: by letting it ask questions, gather clues, and make smart moves step by step under a limited budget.

ArtHOI is a new zero-shot method that makes people and everyday articulated objects (like doors, drawers, and fridges) move together realistically using only a single generated video as guidance.

This paper teaches long-horizon AI agents to remember everything exactly without stuffing their whole memory at once.

Proact-VL is a video-talking AI that knows not only what to say but also when to say it, like a great sports commentator.

DREAM is one model that both understands images (like CLIP) and makes images from text (like top text-to-image models).

Large language models can act unpredictably in sensitive places like schools, hospitals, and customer support, so we need reliable ways to guide how they talk and behave.

MUSE is a new open-source platform that tests how safely AI models behave when you talk to them with text, sound, pictures, and video, not just text.

Reasoning Core is a tool that automatically creates a huge variety of logic and math puzzles, checks every answer with real solvers, and lets you smoothly dial the difficulty up or down.