GenomeOcean Anywhere: Private WebGPU Inference for Genome MoEs

arXiv:2609.35882 · cs.CR, cs.LG, q-bio.GN · Submitted 2026-09-27 · Read on arXiv

cs.CR, cs.LG, q-bio.GN

Submitted: 2026-09-27

Updated: 2026-10-01

Code: https://github.com/ggml-org/llama.cpp

Terminology

• Decode — 41× in this paper · explained in 2 episodes
• Fidelity — 12× in this paper · explained in 2 episodes
• Transformer — 12× in this paper · explained in 2 episodes
• Accuracy — 9× in this paper · explained in 1 episode
• Quantization — 9× in this paper · explained in 12 episodes
• Foundation Model — 6× in this paper · explained in 3 episodes
• Calibration — 4× in this paper · explained in 7 episodes
• Embeddings — 4× in this paper · explained in 2 episodes
• Foundation Models — 4× in this paper · explained in 7 episodes
• Fine-Tuning — 2× in this paper · explained in 5 episodes
• Linear probe — 2× in this paper · explained in 2 episodes
• Mixture of Experts — 2× in this paper · explained in 2 episodes
• Adam — 1× in this paper · explained in 2 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Edge AI — 1× in this paper · explained in 2 episodes
• Edge Devices — 1× in this paper · explained in 2 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Mixture-of-Experts (MoE) — 1× in this paper · explained in 7 episodes
• Mobile Edge Computing — 1× in this paper · explained in 2 episodes
• Model-Agnostic — 1× in this paper · explained in 2 episodes
• Perplexity — 1× in this paper · explained in 3 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Sparsity — 1× in this paper · explained in 4 episodes

Sources

• Mixtral of Experts
• WebLLM: A High-Performance In-Browser LLM Inference Engine
• TensorFlow.js: Machine Learning for the Web and Beyond
• Split learning for health: Distributed deep learning without sharing raw patient data

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs