AI Paper Atlas

Research map / Continual learning / Online and replay-based

How Well Do Unsupervised Learning Algorithms Model Human Real-time and Life-long Learning?

Chengxu Zhuang, Violet Xiang, Yoon Bai, Xiaoxuan Jia, Nicholas Turk-Browne, Kenneth Norman, James J. DiCarlo, Daniel LK Yamins

NeurIPS 2022 · Continual learning › Online and replay-based · approach: Dynamic expansion and isolation · 7 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 SimCLR-style negative sampling to BYOL (BYOLNeg) (ablation)BYOL · Real-time and life-long benchmarksmismatch score / trajectory-averaged Mini-ImageNet accuracyImproved
add SimCLR-style negative sampling to DINO (DINONeg) (ablation)DINO ViT-S · Life-long benchmarktrajectory-averaged Mini-ImageNet accuracyImproved
add SimCLR-style negative sampling to DINO (DINONeg) (ablation)DINO ViT-S · Real-time benchmarkmismatch score to humanNo clear effect
increase within-batch diversity (longer replay window, lower current-memory ratio)all SSL algorithms · Life-long benchmarktrajectory-averaged Mini-ImageNet accuracyImproved
use algorithms without negative sampling (SwAV, BYOL, SimSiam) instead of SimCLR/MoCo v2 in low-diversity streamslife-long continual learning · Life-long benchmark, low diversitytrajectory-averaged Mini-ImageNet accuracyHurt
use MAE instead of contrastive algorithmslife-long continual learning · Life-long benchmark across diversity conditionstrajectory-averaged Mini-ImageNet accuracyNo clear effect
make the current-memory mix ratio extreme: R=1:0 (only current context)SimCLR-More-MLPs · Real-time benchmarkmismatch score to humanHurt
increase current-context share up to R=7:1 from 1:1SimCLR-More-MLPs · Real-time benchmarkmismatch score to humanNo clear effect
lower the current-context share (R=1:7, 1:15, 0:1)SimCLR-More-MLPs · Real-time benchmarkmismatch score to humanHurt
lengthen the current-context replay window to 40m or 80m from 20mSimCLR-More-MLPs · Real-time benchmarkmismatch score to humanHurt
lengthen aggregation time T from 0.2s to 1.0sSimCLR-More-MLPs · Real-time benchmarkmismatch score to human: T=0.2s -> T=1.0sHurt
lengthen aggregation time TSimCLR-More-MLPs · Life-long benchmarktrajectory-averaged Mini-ImageNet accuracyNo clear effect

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