Code: 52475772
We address interpretable representation learning for motion forecasting in self-driving cars. Rather than treating transformers as black boxes, we develop methods to interpret and modify learned representations. We introduce self- ... more
English
33.90 €
RRP: 39.75 €
You save 5.85 €

You get 82 loyalty points
Book synopsis
We address interpretable representation learning for motion forecasting in self-driving cars. Rather than treating transformers as black boxes, we develop methods to interpret and modify learned representations. We introduce self-supervised pre-training with interpretable objectives. Moreover, we probe latent spaces of forecasting models and reveal interpretable features, allowing us to make targeted interventions. Finally, we uncover retrocausal mechanisms, which enable goal-based instructions.
Book details
Book category Books in English Computing & information technology Computer science
33.90 €
EnglishCollection points Bratislava a 12948 dalších
Copyright ©2008-26 najlacnejsie-knihy.sk All rights reservedPrivacyCookies
25900 collection points
Delivery 2.99 €
02/210 210 99 (8-15.30h)Shopping cart ( Empty )