Pith. sign in

Title resolution pending

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.CL 1

years

2019 1

verdicts

CONDITIONAL 1

representative citing papers

Visualizing and Understanding the Effectiveness of BERT

cs.CL · 2019-08-15 · conditional · novelty 5.0

Loss-landscape visualization of BERT fine-tuning indicates pre-training provides a better initialization that yields flatter optima, faster convergence, and lower layers that are more invariant across tasks.

citing papers explorer

Showing 1 of 1 citing paper.

  • Visualizing and Understanding the Effectiveness of BERT cs.CL · 2019-08-15 · conditional · none · ref 5

    Loss-landscape visualization of BERT fine-tuning indicates pre-training provides a better initialization that yields flatter optima, faster convergence, and lower layers that are more invariant across tasks.