AI Paper Atlas

Research map / Optimization / Optimizers

On the Role of Batch Size in Stochastic Conditional Gradient Methods

Rustem Islamov, Roman Machacek, Aurelien Lucchi, Tony Silveti-Falls, Eduard Gorbunov, Volkan Cevher

ICML 2026 · Optimization › Optimizers · approach: SGD momentum and step-size schedules · 0 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
increase batch size from 256 to 4096 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokens, seq len 1024final validation loss: 3.3717 -> 3.6472Hurt
increase batch size from 256 to 512 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokensfinal validation loss: 3.3717 -> 3.3765No clear effect
decrease batch size from 256 to 64 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokensfinal validation loss: 3.3717 -> 3.4554Hurt
increase batch size from 256 to 1024 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokensfinal validation loss: 3.3717 -> 3.4065Hurt
increase batch size from 256 to 2048 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokensfinal validation loss: 3.3717 -> 3.4799Hurt
reduce train sequence length from 1024 to 256 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokens, B=256final validation loss: 3.3717 -> 4.0441Hurt
reduce train sequence length from 1024 to 512 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokens, B=256final validation loss: 3.3717 -> 3.4796Hurt
increase train sequence length from 1024 to 2048 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokens, B=256final validation loss: 3.3717 -> 3.3792No clear effect
increase train sequence length from 1024 to 4096 at beta=3.6e-4 (ablation)Scion 124M NanoGPT · 1.3B tokens, B=256final validation loss: 3.3717 -> 3.4216Hurt
increase train sequence length from 1024 to 2048 at beta=1.2e-4 (ablation)Scion 124M NanoGPT · 1.3B tokens, B=256final validation loss: 3.4587 -> 3.4126Improved
raise Frank-Wolfe stepsize from 1.2e-4 to 2.4e-4 at B=256 (ablation)Scion 124M NanoGPT · 1.3B tokensfinal validation loss: 3.3857 -> 3.3706Improved
raise Frank-Wolfe stepsize from 1.2e-4 to 2.4e-4 at B=4096 (ablation)Scion 124M NanoGPT · 1.3B tokensfinal validation loss: 3.7715 -> 3.6694Improved

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