WEZZEL NEWS

Object-Uni: A Unified Model for Object-Centric Spatial Understanding and Controllable Generation

Researchers have proposed a new AI model called Object-Uni that can understand and manipulate the sp...

XTC: Head-Aware Sampling by Excluding Top Choices

Researchers have developed a new decoding operator called XTC (Exclude Top Choices) for autoregressi...

Don't Repeat Yourself: Stopping Verbatim Loops at Sampling Time

Researchers propose a new method called Don't Repeat Yourself (DRY) to prevent large language models...

TRACE: A Self-Evolving Skill Bank for Consistent, Limit-Aware LLM Agents

A new system called TRACE aims to improve the reliability of large language models (LLMs) in real-wo...

Triplet2Track: A Hierarchical System with Object-Centric Representations for Reliable Long-Horizon Manipulation

Researchers have developed a system called Triplet2Track that enables robots to ...

SDoH-Aware Narrative Anchoring Bias in Medical LLMs for Trustworthy Clinical Decision Support

Researchers have identified a potential issue with medical large language models...

Fairness-Aware Mixture-of-Experts via Subgroup Reweighting and Gate Regularization

Researchers have developed a new framework to address performance disparities in...

Minimal Local Simulation Foundations for LLM- and VLM-Driven Agents in 2D and 3D Environments

Researchers have created two minimal simulation foundations for large language m...