CalOn: a 0.6B unified encoder tackles entity, relation and event extraction across schemas
Yang Wang · hf · 2026-10-16
University of Manchester with Mayo Clinic and others released CalOn, a compact 0.6B unified information extraction encoder.
- Motivation: IE models transfer poorly to new domains and annotation guidelines; supporting user-specified types alone doesn't ensure accuracy on new datasets
- Method: jointly encodes text with type definitions, performs entity, relation, and event extraction in one framework, and applies shared probabilistic reasoning over combined type evidence for calibrated typed decisions
- One model serves multiple tasks and schemas without per-guideline training
Practical for teams deploying extraction pipelines across domains.
More from Research
- Is math just a language? A debate on why LLMs are good at mathematics — LucaAmb · 2026-10-16
- Alibaba open-sources OpenGCUA, a full-stack generalist computer-use agent; +19.6 points on OSWorld-v2 across 8 models — Tongyi-MAI · 2026-10-16
- CoDA co-evolves simulation and robot policies via shared failure memory, lifting OOD success from 5.3% to 27.6% — PLAN-Lab · 2026-10-16
- Cambridge NLIP seminar: Danae Sanchez Villegas on multimodal reasoning beyond final answers — nikaletras · 2026-10-16
- DKGI paper: agents write their own tools to search knowledge graphs, beating baselines by 6.7 Hit@1 — yuz9yuz · 2026-10-16
- ETH Zurich's LFD lets a frozen LLM teach drug-design generators natural-language goals — rohanpaul_ai · 2026-10-16