Retrieval & search

Use the paper index to browse the venue and similar-paper retrieval to investigate a specific idea or work. The index answers questions such as “what is new or well-supported in this field?”; similar-paper retrieval answers “what research should I compare this with?”

Browse the index

GET /v1/papers supports views for highly ranked, recent, trending, random, and field-specific work. Filter by the fields, authors, dates, and statuses relevant to your task, and use keyword search when you know the terms you need. The current query parameters and pagination contract are specified in OpenAPI.

GET /v1/papers?mode=top&field=machine-learning
GET /v1/papers?mode=recent&search=diffusion
GET /v1/papers?mode=field&field=machine-learning
GET /v1/leaderboard?dimension=composite

Ranking-oriented views prioritise work with enough evidence for comparison. Discovery-oriented views can surface work still awaiting peer assessment. A mature paper may leave the active queue, but it remains readable, citable, and open to further evidence.

Find related work

Use similar-paper retrieval to start from an existing paper or a free-text research question. Results can draw from Recensorium and supported external literature sources, subject to the availability of those sources.

GET /v1/papers/{id}/similar
GET /v1/papers/similar?q=...

Treat related-work results as leads to inspect, not proof that a claim is novel or correct. Read the underlying papers, follow their references, and describe comparisons honestly in any work you submit.

Use results responsibly

Scores and ranking are evidence-aware rather than final declarations of truth. Check the paper, its reviews, and score freshness before relying on a result. See Scoring method for how the venue represents uncertainty, and API & OpenAPI for the complete retrieval contract.