AI Paper Atlas

Research map / Graph learning / Graph algorithms and clustering

Learning to Navigate Wikipedia by Taking Random Walks

Manzil Zaheer, Kenneth Marino, Will Sussman Grathwohl, John Schultz, Wendy Shang, Sheila Babayan, Arun Ahuja, Ishita Dasgupta, Christine Kaeser-Chen, Rob Fergus

NeurIPS 2022 · Graph learning › Graph algorithms and clustering · approach: Continuous DAG structure learning · 1 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
train on random forward trajectories instead of reverse trajectories (ablation)RFBC · 200k graph forward navigation 20-stepsuccess rate (%): 0.2 -> 67.5Improved
train on shortest-path trajectories instead of random forward (ablation)RFBC · 200k graph forward navigation 20-stepsuccess rate (%): 67.5 -> 84.6Improved
use random features instead of RoBERTa embeddings (ablation)RFBC · 200k graph navigation 5-stepsuccess rate (%): 85.3 -> 17.4Hurt
replace behavioral cloning with RL (IMPALA) (ablation)RFBC · 200k graph navigation 20-stepsuccess rate (%): 67.5 -> 41.6Hurt
replace behavioral cloning with RL (ablation)RFBC · 200k graph sentence search 10-stepsuccess rate (%): 47.6 -> 1.5Hurt
fine-tune RFBC policy with RL (ablation)RFBC · 200k graph navigation 20-stepsuccess rate (%): 67.5 -> 68.1No clear effect
train embeddings on navigation instead of fixed RoBERTa (ablation)feed-forward policy · 38M graph sentence search 20-stepsuccess rate (%): 77.5 -> 90.2Improved
train embeddings on navigation instead of fixed RoBERTa (ablation)feed-forward policy · 38M graph navigation 20-stepsuccess rate (%): 71.5 -> 89.8Improved
use 4-layer Transformer policy instead of feed-forward (ablation)learned embeddings · 38M graph navigation 20-stepsuccess rate (%): 89.8 -> 92.2Improved
use 4-layer Transformer policy instead of feed-forward (ablation)learned embeddings · 38M graph sentence search 20-stepsuccess rate (%): 90.2 -> 79.7Hurt
scale training from 200k graph to full 38M graphRFBC · sentence search 20-stepsuccess rate (%): 34.6 -> 90.2Improved
add RFBC navigation to BM25 + TF-IDF (ablation)BM25 + TF-IDF · FEVER devF1@5: 0.36 -> 0.46Improved

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