Benchmark for evaluating end-to-end web-app generation capabilities of LLM-based agents.

Model checkpoints and synthetic datasets for omnimodal world models.

A deterministic, sparsity-aware evaluation algorithm for P-Complete queries over inverted indices.

An open-source Python package to generate graph alignment datasets and benchmark GNN architectures.

A decision-making framework for robust robotic manipulation in high uncertainty environments.

A framework for autonomous skill acquisition in steerable Vision-Language-Action (VLA) models.

Unified vision-language and grounding framework for interpretable risk understanding in autonomous driving.

A multi-agent framework for sanitizing retrieved content in RAG via semantic rewriting.

Causal model generalization framework for fragmented biomedical data.

Open measurement tool for quantifying tokenization premiums in African languages.

Mother of All BCI Benchmarks framework for evaluating brain-computer interface decoders.

A local-first, agentic AI operating system utilizing markdown for memory and orchestration.

A world model for predicting environment states to enhance AI agent training and performance.

RL-based trigger control for the Large Hadron Collider

AI agent harness that runs entirely in your browser

Cross-view Aligned Evidence-driven Multimodal Reinforcement Learning for 3D Tumor Analysis

A Ruby framework for interacting with major AI providers.

A system for computing apothem-optimal robustness certifications for neural networks.

Frequency-guided feature representation learner for efficient small object detection.

A dual-branch spiking neural network for energy-efficient neuromorphic speech enhancement.