AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Locality Sensitive Teaching

Zhaozhuo Xu, Beidi Chen, Chaojian Li, Weiyang Liu, Le Song, Yingyan Lin, Anshumali Shrivastava

NeurIPS 2021 · Learning theory and algorithms › Prediction and decision losses · approach: Data attribution and recourse · 2 citations (OpenAlex)

Read the paper

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
replace IMT exact greedy selection with LSH-based sampling (LST)IMT · sliceseconds per iteration: 404.95 -> 0.196Improved
replace IMT with LSTIMT · 4 regression datasetstime to same train loss: 2000x fasterImproved
replace IMT with LSTIMT · abalone on TX2speedup: 425.12×Improved
replace IMT with LSTIMT · abalone on TX2energy savings: 99.76%Improved
replace IMT with LSTIMT · space ga on TX2speedup: 149.07×Improved
replace IMT with LSTIMT · mg on TX2speedup: 117.20×Improved
replace IMT with LSTIMT · synthetic classification on TX2speedup: 306.12×Improved
replace IMT with LSTIMT · regression datasetsiteration-wise train and test lossNo clear effect
replace stochastic SGD teaching with LST or IMTlearner · regression datasetsiterations to convergeImproved
replace IMT with LST on sliceIMT · sliceiteration-wise loss: about 30% LST instances outperform IMTMixed
replace GTA with AIPS (randomized sampling) in toy example (ablation)GTA · toy logistic regressiontrain loss per iterationImproved
run LST on TX2 instead of serverLST · regression datasetsiteration-wise convergenceNo 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