AI Paper Atlas

Research map / Graph learning / GNN architectures

Learning Multi-Task Communication with Message Passing for Sequence Learning

Pengfei Liu, Jie Fu, Yue Dong, Xipeng Qiu, Jackie Chi Kit Cheung

AAAI 2019 · Graph learning › GNN architectures · approach: Expressive equivariant message passing · 19 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 learned direct inter-task message passing (DC-MTL) over shared-private sharingSP-MTL · 16 text classification taskserror rate (%): 13.9 -> 12.5Improved
replace shared-private layers with mailbox virtual node and selective retrieval (IC-MTL)SP-MTL · 16 text classification taskserror rate (%): 13.9 -> 12.7Improved
learn multi-task communication with message passing instead of fully shared LSTMFS-MTL · 16 text classification taskserror rate (%): 15.3 -> 12.5Improved
transfer frozen shared layer from IC-MTL with attention retrieval instead of SP-MTL shared layerSP-MTL · 16 text tasks leave-one-task-out transfererror rate (%): 14.3 -> 13.1Improved
multi-task transfer (IC-MTL) vs single-task trainingsingle-task LSTMs · 16 text tasks leave-one-task-out transfererror rate (%): 18.0 -> 13.1Improved
use IC-MTL instead of best baseline FS-MTLFS-MTL · CoNLL 2003 NERF1: 90.94 -> 91.47Improved
use IC-MTL instead of FS-MTLFS-MTL · CoNLL 2000 chunkingF1: 95.41 -> 95.61Improved
use IC-MTL instead of FS-MTLFS-MTL · PTB POS taggingaccuracy: 97.55 -> 97.69Improved
multi-task training with some baselines (FS-MTL, MT-CNN) vs single task on individual review tasksFS-MTL · Apparel, Video, Magazines (text classification)error rate (%): Apparel 15.7 -> 16.3Mixed

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