A single transformer training recipe, poly-autoregressive prediction, improves multi-agent ego forecasting over autoregressive baselines on three distinct tasks.
Argoverse: 3d track- ing and forecasting with rich maps
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Poly-Autoregressive Prediction for Modeling Interactions
A single transformer training recipe, poly-autoregressive prediction, improves multi-agent ego forecasting over autoregressive baselines on three distinct tasks.