Code and audit artifacts for auditable repair of scientific reasoning graph extraction.

Implementation of the Attention-Guided Memory Refinement framework for agent memory.

Open-source SDK and CLI for structured memory in AI companions, focusing on emotional continuity.

Implementation of Reward-Driven LLM Agent Workflows using POMDP routing and self-correction.

Implementation and checkpoints for Continuous Chain-of-Thought models using direct supervision.

Source code for a hybrid multi-agent LLM system for conversational depression screening.

A reactive graph engine in Julia for neural network circuit discovery and interventions.

Open-source spatial index of displaced cultural artifacts

Self-hosted SEO monitoring dashboard with GSC integration and MCP server for AI agents.

Comprehensive documentation for enabling LLM fine-tuning and inference on AMD hardware.

A cryptography-first networking stack designed to work without a central internet infrastructure.

An open-source model designed to detect and mitigate indirect prompt injection attacks in LLM agents.

A structure-agnostic benchmark for evaluating the ground-up generation of CLI tools by LLM agents.

Implementation and results of LLM integration for Task and Motion Planning with PDDLStream.

Knowledge orchestration pipeline for converting raw documents into scientific knowledge graphs.

An LLM fine-tuned via RL-Struct for high structural accuracy in JSON generation.

A streamlined multi-scale feature fusion decoder for efficient ViT-based multi-task dense prediction.

A hierarchical world model for driving with code, demo, and checkpoints.

A noise-guided optimization framework for realistic autoregressive video generation.

Adaptive runtime for guarding against memory traps in robot navigation.