Pith. sign in

REVIEW 7 cited by

Centaur: Robust End-to-End Autonomous Driving with Test-Time Training

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2503.11650 v1 pith:F23C3YXQ submitted 2025-03-14 cs.RO cs.AIcs.CVcs.LG

classification cs.ROcs.AIcs.CVcs.LG
keywords test-timetrainingcentaurclustercostentropyplanneruncertainty
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

How can we rely on an end-to-end autonomous vehicle's complex decision-making system during deployment? One common solution is to have a ``fallback layer'' that checks the planned trajectory for rule violations and replaces it with a pre-defined safe action if necessary. Another approach involves adjusting the planner's decisions to minimize a pre-defined ``cost function'' using additional system predictions such as road layouts and detected obstacles. However, these pre-programmed rules or cost functions cannot learn and improve with new training data, often resulting in overly conservative behaviors. In this work, we propose Centaur (Cluster Entropy for Test-time trAining using Uncertainty) which updates a planner's behavior via test-time training, without relying on hand-engineered rules or cost functions. Instead, we measure and minimize the uncertainty in the planner's decisions. For this, we develop a novel uncertainty measure, called Cluster Entropy, which is simple, interpretable, and compatible with state-of-the-art planning algorithms. Using data collected at prior test-time time-steps, we perform an update to the model's parameters using a gradient that minimizes the Cluster Entropy. With only this sole gradient update prior to inference, Centaur exhibits significant improvements, ranking first on the navtest leaderboard with notable gains in safety-critical metrics such as time to collision. To provide detailed insights on a per-scenario basis, we also introduce navsafe, a challenging new benchmark, which highlights previously undiscovered failure modes of driving models.

Discussion (0). Sign in to comment.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Test-Time Coverage: Test-Conditioned Data Curation for Deployment-Aware Learning

    cs.AI 2026-07 conditional novelty 6.0 of 10

    TTCov curates training data for deployment by building an LLM-generated atomic-proposition atlas of the test distribution and greedily selecting clips that match it.

  2. Zero-Human Demonstration End-to-end Autonomous Driving with Trajectory Scorer

    cs.RO 2025-10 conditional novelty 6.0 of 10

    A reward-only offline RL method for trajectory planning in end-to-end autonomous driving achieves state-of-the-art on Navhard and competitive closed-loop HUGSIM performance without imitation learning.

  3. Reinforced Refinement with Self-Aware Expansion for End-to-End Autonomous Driving

    cs.RO 2025-06 reject novelty 6.0 of 10

    R2SE refines pretrained end-to-end driving policies on hard cases via residual LoRA reinforcement learning and switches between specialist and generalist policies using GPD-based uncertainty.

  4. ETA: Efficiency through Thinking Ahead, A Dual Approach to Self-Driving with Large Models

    cs.CV 2025-06 conditional novelty 6.0 of 10

    An asynchronous dual-system architecture forecasts large-model features into the current frame and adds a small-model update to drive in near real time, scoring 69.53 on Bench2Drive at 50 ms.

  5. HAD: Combining Hierarchical Diffusion with Metric-Decoupled RL for End-to-End Driving

    cs.RO 2026-04 conditional novelty 5.5 of 10

    Hierarchical diffusion plus polar structure-preserving expansion and metric-decoupled RL yields SOTA open- and closed-loop planning scores on NAVSIM and HUGSIM.

  6. Generalized Trajectory Scoring for End-to-end Multimodal Planning

    cs.RO 2025-06 conditional novelty 5.0 of 10

    GTRS combines super-dense vocabulary training, dropout, sensor augmentation, and diffusion proposals to reach 49.4 EPDMS on the Navhard benchmark, approaching the privileged PDM-Closed method.

  7. HMAD: Advancing E2E Driving with Anchored Offset Proposals and Simulation-Supervised Multi-target Scoring

    cs.CV 2025-05 conditional novelty 4.0 of 10

    HMAD integrates BEVFormer, DiffusionDrive-style anchor offsets, and a Hydra-MDP-style scoring network to achieve 65.94 EPDMS on the NAVSIM warmup benchmark and 44.5% on the CVPR 2025 private test set.

Pith tools