Research memory for the HSSM ecosystem

HSSMCortex turns reviewed research summaries and paper text into a searchable knowledge graph and vector store. It supports contributors who need to find prior work, trace relationships between methods, or maintain the ecosystem's curated research record.

Run your first query Understand the ingestion path

Early-stage operator tool

HSSMCortex is a capability repository, not a released end-user package. Its query and ingestion commands require a locally populated Cognee store; ingestion can also require private Drive access and an Anthropic API key. The documentation site itself is static and builds without those services, credentials, or the HSSMCortex runtime.

Choose a starting point

Goal	Start here
Search an existing local knowledge store	Run your first query
Refresh the store from reviewed summaries	Ingest reviewed content
Operate or contribute to Cortex	How-to guides
Understand design and provenance	Explanations
Look up commands and data contracts	Reference

What Cortex owns

• Paper metadata and reviewed summaries under papers/.
• Research taxonomies and curated guidance that have passed repository review.
• The local Cognee ingestion, query, graph, and entity-resolution tooling.
• Capability content that the Spine can distribute to coding agents.

Cortex does not own package APIs, scientific usage guidance, network training, or pipeline operation. Follow the owning site for those subjects.

Native documentation sites

• HSSM — Bayesian inference with sequential sampling models.
• ssm-simulators — simulation engines and model definitions.
• LANfactory — likelihood-approximation network training.
• LAN pipeline — data-generation and training operations.
• HSSMCortex — research knowledge and capabilities.
• HSSMSpine — ecosystem coordination and shared contracts.