Visual framework for building multi-agent AI systems.

C++ DSP modules for Max and Ableton Live

Variable-length tokenizer via learnable global merging for Diffusion Transformers.

Backbone for 2D and 3D thin-structure segmentation in medical imaging.

A module for scale-calibrated camera-motion transfer using a gauge-invariant parallax number.

Dataset for evaluating cultural knowledge in dance across 12 languages.

Visual analytics tool for evaluating LLM bias through aggregated stochastic generation paths.

An agentic reward framework that evaluates behaviors to provide robust signals and mitigate reward hacking.

Embodiment-aware co-speech motion generation framework for humanoid robots.

A natural language interface for querying global explanations of black-box image classifiers.

Stereotyped Object Diagnostic Audit framework for measuring bias in text-to-image models.

A training-free framework for hierarchical multi-agent scientific poster generation.

Multi-agent LLM system for scaling feedback on software architecture deliverables.

A recommender system using sparse autoencoders to construct fine-grained interpretable intents priors.

A visual framework for building multi-agent AI pipelines and RAG applications.

A model-centric retriever attack framework for malicious knowledge injection in RAG systems.

A proposed markup language designed for AI agent-optimized content and human supervision tiers.

A research harness for externalizing and validating AI scientific research processes.

A unified framework for Dual-Path Spatial Reasoning in Spatial Vision Language Models.

EHR-grounded evaluation framework for assessing clinical utility of LLMs in medical workflows.