Unleashing VLA Potentials in Autonomous Driving via Explicit Learning from Failures

TL;DR

This paper introduces ELF-VLA, a framework that enhances Vision-Language-Action models for autonomous driving by using explicit failure diagnostics to guide reinforcement learning, significantly improving performance in challenging scenarios.

Contribution

We propose ELF-VLA, which incorporates structured failure feedback into RL to address exploration limitations and improve autonomous driving performance.

Findings

Achieves state-of-the-art results on NAVSIM benchmark.

Effectively identifies specific failure modes in driving scenarios.

Enables VLA models to solve critical long-tail scenarios.

Abstract

Vision-Language-Action (VLA) models for autonomous driving often hit a performance plateau during Reinforcement Learning (RL) optimization. This stagnation arises from exploration capabilities constrained by previous Supervised Fine-Tuning (SFT), leading to persistent failures in long-tail scenarios. In these critical situations, all explored actions yield a zero-value driving score. This information-sparse reward signals a failure, yet fails to identify its root cause -- whether it is due to incorrect planning, flawed reasoning, or poor trajectory execution. To address this limitation, we propose VLA with Explicit Learning from Failures (ELF-VLA), a framework that augments RL with structured diagnostic feedback. Instead of relying on a vague scalar reward, our method produces detailed, interpretable reports that identify the specific failure mode. The VLA policy then leverages this explicit feedback to generate a Feedback-Guided Refinement. By injecting these corrected, high-reward samples back into the RL training batch, our approach provides a targeted gradient, which enables the policy to solve critical scenarios that unguided exploration cannot. Extensive experiments demonstrate that our method unlocks the latent capabilities of VLA models, achieving state-of-the-art (SOTA) performance on the public NAVSIM benchmark for overall PDMS, EPDMS score and high-level planning accuracy.