TERRA: Tissue World Model Pretrained on 112M Cells
mo_lotfollahi · x · 2026-08-06
Researchers introduced TERRA, a world model for human tissues. The project spanned 1.5 years, involving a massive data-generation and modeling effort pre-trained on 112M cells from spatial transcriptomics.
- Data Scale: Built on one of the largest human spatial transcriptomics corpora to date, it spans 20 tissues across development, health, and 26 disease conditions, with two-thirds newly generated in-house.
- Methodology: Inspired by Yann LeCun's JEPA approach, TERRA learns via prediction in embedding space, aiming to build a universal representation capturing the multi-scale logic from genes to cells and their microenvironments.
Related event: TERRA: Tissue World Model Pretrained on 112M Cells(3 posts)→
More from Research
- ARCHead: New LLM Output Head Quantization Method Substantially Reduces Storage with Minimal Loss — Şuayp Talha Kocabay · 2026-08-06
- CALVER: Causal Verification Method for LLM Reasoning Breaks Self-Consistency Bottleneck — JerzakLabs · 2026-08-06
- LegalPincite: Multi-level Legal IR Dataset Addresses Paragraph-Level Citation Gaps — Theresia Veronika Rampisela · 2026-08-06
- TREC 2026 RAG Track Opens: Featuring New Nvidia ClimbMix Corpus — lintool · 2026-08-06
- New paper bridges information nuggets and side-by-side comparisons using LMSYS data for complex LLM response evaluation — lintool · 2026-08-06
- TreeAdapter Integrates 10,000+ LoRAs into a Single 4B Model System — bdsqlsz · 2026-08-06