AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

scDataset: Scalable Data Loading for Deep Learning on Large-Scale Single-Cell Omics

Davide D'Ascenzo, Sebastiano Cultrera di Montesano

ICML 2026 · Architectures and efficiency › RNNs and normalization · approach: Distributed LLM inference systems · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add block sampling with batched fetching to replace per-cell random samplingAnnData loading on Tahoe-100M · Tahoe-100Mthroughput: ~20 samples/sec -> over 204x higherImproved
increase fetch factor from 1 to 1024 (ablation)scDataset streaming · Tahoe-100M streamingthroughput: 345 -> 5263 samples/sec (over 15x)Improved
add batched fetching (f=256) with block size 16 (ablation)scDataset block sampling · Tahoe-100M minibatch entropyplate label entropy: 1.76 -> 3.61Improved
increase block size to 64 or more without batched fetching (ablation)scDataset · Tahoe-100M minibatch entropyplate label entropy: entropy near zero at b>=64, f=1Hurt
use sequential streaming instead of random or block shuffling (ablation)drug classification · Tahoe-100M drug classification, linearmacro F1: 0.030 -> 0.002Hurt
use block shuffling b=16 f=256 instead of random sampling (ablation)linear and MLP classifiers · Tahoe-100M classification tasksmacro F1: MoA broad linear 0.270 -> 0.264No clear effect
add a 16,384-cell shuffle buffer to streaming (ablation)streaming loader · Tahoe-100M classification tasksmacro F1: MoA broad MLP 0.185 -> 0.201No clear effect
change sampling strategy on cell line classification (ablation)linear and MLP classifiers · Tahoe-100M cell line classificationmacro F1: linear 0.935-0.937; MLP 0.942No clear effect
use streaming instead of block shuffling (ablation)training dynamics · Tahoe-100M training losstraining lossHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy