Research intelligence, not just paper summaries.
Papers scored by importance and readability · Authors & labs tracked by momentum · Research gaps surfaced daily · Creator-ready content generated automatically
Every day we scan 50,000+ papers across arXiv CoRR and 15+ top conference proceedings. Here's how we narrow it down to the 1,000 worth your time.
Institution & Author Prestige Scoring
Papers from top labs (OpenAI, DeepMind, Google Research, MIT, Stanford…) and renowned researchers now get a scoring boost. Ambiguous names require institution corroboration to avoid false positives.
Conference Papers Database (50K+)
Permanent conference paper store covering NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, ACL, EMNLP, NAACL and more. Conference papers never expire and are ranked separately from arXiv preprints.
CoRR-Only arXiv Feed
Daily arXiv updates now strictly restricted to CoRR (Computing Research Repository) — all 19 cs.* categories. Non-CS papers from physics, statistics, or engineering are filtered out.
Topics Page — Accordion + Difficulty Filters
80 research topics now collapsed by default with click-to-expand. Filter by difficulty (Beginner → Frontier) or search by name. Advanced topics like RLHF, DPO, GRPO, MoE, SSMs correctly classified as L3.
More Conference Venues
AAMAS, ICAPS, KDD, ICDM, RecSys, SIGIR, ICSE — waiting on Semantic Scholar API key for bulk access.