AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving

Ziying Song, Lin Liu, Hongyu Pan, Bencheng Liao, Mingzhe Guo, Lei Yang, Yongchang Zhang, Shaoqing Xu, Caiyan Jia, Yadan Luo

TPAMI 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors · 25 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 map and agent conditioning to the diffusion generator (ablation)SparseDrive · Bench2DriveDriving Score: 44.54 -> 45.34Improved
add reference GT guidance (ablation)SparseDrive · Bench2DriveDriving Score: 44.54 -> 47.12Improved
add RL with diversity and safety rewards (ablation)SparseDrive + PADG · Bench2DriveDriving Score: 47.12 -> 49.21Improved
add map and agent conditioning (ablation)TransFuser mmt · NAVSIM v1PDMS: 85.1 -> 86.3Improved
add GRPO RL on top of reference GTs and match loss (ablation)DIVER PADG · NAVSIM v1PDMS: 86.7 -> 88.3Improved
use GRPO instead of PPO (ablation)DIVER · NAVSIM v1PDMS: 87.9 -> 88.3Improved
use TrajPooler instead of no image features (ablation)DIVER · NAVSIM v1PDMS: 87.7 -> 88.3Improved
use TrajPooler instead of full image features (ablation)DIVER · NAVSIM v1PDMS: 87.9 -> 88.3Improved
add traffic-compliance reward weight 0.1 (ablation)DIVER · Bench2DriveDriving Score and Comfortness: 49.21 -> 48.71; 54.72 -> 56.10Mixed
add PDMS reward weight 0.1 (ablation)DIVER · NAVSIM v1PDMS: 88.3 -> 88.6No clear effect
apply DIVER to SparseDriveSparseDrive · nuScenesavg diversity: 0.13 -> 0.21Improved
apply DIVER to SparseDriveSparseDrive · Adv-nuSccollision rate 3s: 2.430 -> 1.798Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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