# Lacuna, the open map of machine learning research by Tiptree Systems Canonical: https://tiptreesystems.com/lacuna A living library of machine learning research: more than two million papers distilled into research directions, research proposals, and author pages. Open to anyone, no account needed. Open commons No account needed Built by Tiptree Systems [Explore Lacuna](https://lacuna.tiptreesystems.com) [Read the paper](https://tiptreesystems.com/research/lacuna) ## What the map holds Lacuna reads the papers it indexes and builds four layers on top. Every generated item keeps links to the primary sources that support it, so wherever you land, [the trail back to the original is one click](https://docs.tiptreesystems.com/trust-provenance). [2.2M+ papers](https://lacuna.tiptreesystems.com) [38k+ research directions](https://lacuna.tiptreesystems.com/directions) [210k+ authors](https://lacuna.tiptreesystems.com/authors) ### Paper pages Markdown expositions of each paper’s core ideas, some with the key figures extracted and kept. ### Concept elements One- or two-sentence statements of a core idea, method, limitation, or observation, each linked to its papers. ### Research directions Two-page syntheses of concept clusters: a recurring problem, method family, or opportunity area. ### Research proposals Generated hypotheses, sampled from directions and their supporting papers. ## How the map is built Lacuna distills the literature into searchable research directions and new proposals, connected to the researchers advancing them. Lacuna does not rely on citations, but on the semantic content of papers to determine relatedness. If you work in ML, your research is there. The pipeline harvests and reconciles scholarly records from OpenReview, arXiv, venue open-access pages, OpenAlex, and DBLP, and it establishes stable identifiers before generating anything, so trails cannot dangle. Author pages are anchored on author-maintained OpenReview records rather than external author IDs that conflate researchers who share a name. The research proposals are not free-form LLM ideation. They are produced with a published method we call [alien sampling](https://tiptreesystems.com/research/alien-space-of-science): combinations of ideas that are plausible under the structure of existing knowledge but unlikely under the distribution of existing researchers. ## Use it from anywhere The same map for people and for agents: a page on the web, the same page as text, and a protocol a coding agent can call. ### On the web Browse papers, research directions, authors, venues and institutions. No account, no paywall. [Open Lacuna](https://lacuna.tiptreesystems.com) ### As Markdown Read any page as text by adding `/md` to its address, for reading in a terminal or feeding to a model. [See the Markdown interface](https://lacuna.tiptreesystems.com/md) ### From an agent Agents connect through the Model Context Protocol server. [Althea](https://tiptreesystems.com/althea), the research assistant by Tiptree Systems, works over Lacuna and over the parts of your research the commons cannot see. [Connect an agent](https://docs.tiptreesystems.com/lacuna-machine) ## About the name *lacuna*, noun, Latin: a gap. Lacuna by Tiptree Systems exists to tend and grow the open research culture that made modern AI possible, and to point at the gaps in our collective understanding. It is built and run by Tiptree Systems, an R&D company in Montreal and Berlin, and benchmarked in a public preprint presented at the ICML Workshop on AI for Research. [Explore Lacuna](https://lacuna.tiptreesystems.com) [Meet Althea, the research assistant](https://tiptreesystems.com/althea)