ManeuverFlow: Maneuver Hypothesis Generation with Future Evidence for Trajectory Evaluation and Refinement in End-to-End Autonomous Driving

arXiv 2026 32.6 method

TLDR

ManeuverFlow generates diverse trajectories for autonomous driving via maneuver-routed mixture-of-experts and evaluates/refines them using future features from a world model, achieving strong NAVSIM and WorldEngine benchmark results.

Reasoning

The paper addresses important limitations in end-to-end planning, such as mode collapse and scaling of candidate sets, and demonstrates strong empirical results on multiple benchmarks. However, the abstract gives limited detail on the world model's design and the evaluations are primarily simulation-based, so the generalizability to real-world deployment is not fully established.

Read-first score

Read-first score 32.6, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 24.

Recency 5%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Reproducibility 15%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=True; dataset=False; markers=code,github

Methodology quality 15%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark,evaluation

Topical relevance 25%

34.3

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Graph prestige 15%

5.7

PageRank-style prestige over the local citation graph built from available OpenAlex references. local_pagerank=0.0016

Citation impact 15%

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. cited_by_count=0

Citation velocity 10%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FrontierReproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 351.

Keyword Scores

world model

world dynamics prediction

generative world model

interactive world model

world simulator

video world model

model-based reinforcement learning world model

Resources

Paper Code