Embeddings × Developer Tools

OlmoEarth v1.3 adds global embeddings

OlmoEarth v1.3 adds global embeddings

✎ Story body

OlmoEarth v1.3 released planet-scale embeddings, and with this week's embedding papers it hints at embeddings spreading beyond language.

What happened

A Hugging Face blog post presented OlmoEarth v1.3 as state-of-the-art embeddings for the whole planet. Four arXiv papers landed the same day, including the pplx-embed-v2-late technical report and work on RAG compositionality.

Why it matters

It suggests embeddings are becoming shared building blocks in non-language domains such as geospatial data. The details and performance level of the release are not confirmed from this material, and the accompanying papers are not uniform in theme. Note that this is not yet an established trend.

What to watch

Real-world uptake and downstream-task evaluations of OlmoEarth, and any design overlap with language-side embedding methods such as late interaction.

▲ Official & Press
Official

OlmoEarth v1.3: State-of-the-art embeddings for the whole planet

Hugging Face Blog ・ 2026-10-22 ・ 📌

OlmoEarth v1.3 released with state-of-the-art whole-planet embeddings

Academic (arxiv etc.) 10 ▾
Academic

Loopholing All the Way: Language Modeling with pure Self-Conditioning

arXiv cs.LG (Machine Learning) ・ 2026-10-21

Academic

Code Comprehensor: Hierarchical Documentation and Cross-Modal Retrieval for Large Repositories

arXiv cs.AI (Artificial Intelligence) ・ 2026-10-21

Academic

EPIC: Rethinking Trust Regions for Autoregressive LLM Policies via Exact-KL Prefix Importance-Ratio Clipping

arXiv cs.LG (Machine Learning) ・ 2026-10-21

Academic

Post-hoc Epistemic Uncertainty for Vision-Language-Action Policies

arXiv cs.LG (Machine Learning) ・ 2026-10-21

Academic

Retrieval Augmented Generation Compositionality

arXiv cs.CL (Computation and Language) ・ 2026-10-21

Academic

Rethinking Post-hoc Debiasing of Large Pretrained Models

arXiv cs.LG (Machine Learning) ・ 2026-10-21

Academic

SPARE: Sparsity Allocation via Routing Entropy for Efficient MoEs

arXiv cs.CL (Computation and Language) ・ 2026-10-21

Academic

LLMs as survey respondents: Comparison to simple semantic baselines

arXiv cs.CL (Computation and Language) ・ 2026-10-21

Academic

pplx-embed-v2-late Technical Report

arXiv cs.CL (Computation and Language) ・ 2026-10-21

Academic

A Strong MoE Is All You Need for Process Supervision

arXiv cs.CL (Computation and Language) ・ 2026-10-21

← Story Archive