Dual-channel hypergraph learning framework for node importance estimation in heterogeneous knowledge graphs.

Controlled human-robot co-training framework for world action models.

Theory-agnostic framework for personality recognition using prototypical networks

An open-source, privacy-focused firewall for securing LLM interactions via browser extension and proxy.

A lightweight adaptive multi-rhythm network for long-term time series forecasting.

A graph-based framework for quantifying uncertainty, coherence, and robustness in LLM logic.

WASM-based FET-level circuit simulation and visualization tool.

A curated repository of methods, datasets, and benchmarks for multimodal unlearning.

Recursive Consolidation of Low-Rank Adapters for continual LLM fine-tuning.

A 5-million-sample bilingual document corpus for parsing, layout analysis, and VQA.

A full operating system written in Rust, focusing on safety and concurrency.

A structured belief-store proxy that resolves scope and retrieval before inference to improve LLM memory.

Dynamic Audio-Driven Semantic Chunking for efficient omnimodal token compression.

Implementation of the Hierarchical Mixture-of-Experts for generalist vision-language-action policies.

A general-purpose, fully implicit soft-body simulator for handling soft dynamics and frictional contact.

Evaluation framework and benchmarks for VLM reasoning about object properties.

A free way to run several Claude Code agents in one window.

Interactive visual map of millions of research papers.

Open source product analytics platform

Open-source benchmark for evaluating AI chatbot safety in mental health contexts.