# Cross-Stage Coherence in Hierarchical Driving VQA: Explicit Baselines and Learned Gated Context Projectors arXiv: https://arxiv.org/abs/2604.22560 ยท HTML: https://mlsanity.com/paper/2604.22560/ ## Key facts - No citing papers in the map yet (data through 14 Oct 2026). - Placed in VLMs for autonomous driving, one of 9 directions in End-to-end autonomous driving. - Builds on 32 of its 37 references that are in the map. - First posted on arXiv 24 Apr 2026. ## Placed in - End-to-end autonomous driving / [VLMs for autonomous driving](https://mlsanity.com/driving/vlms-for-autonomous-driving/) ## Builds on (most cited first, 30 of 32) - [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021) - [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019) - [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023) - [CIDEr: Consensus-based Image Description Evaluation](https://mlsanity.com/paper/1411.5726/) (2014) - [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025) - [Training Large Language Models to Reason in a Continuous Latent Space](https://mlsanity.com/paper/2412.06769/) (2024) - [DriveLM: Driving with Graph Visual Question Answering](https://mlsanity.com/paper/2312.14150/) (2023) - [DriveVLM: The Convergence of Autonomous Driving and Large Vision-Language Models](https://mlsanity.com/paper/2402.12289/) (2024) - [LMDrive: Closed-Loop End-to-End Driving with Large Language Models](https://mlsanity.com/paper/2312.07488/) (2023) - [CODI: Compressing Chain-of-Thought into Continuous Space via Self-Distillation](https://mlsanity.com/paper/2502.21074/) (2025) - [OpenDriveVLA: Towards End-to-end Autonomous Driving with Large Vision Language Action Model](https://mlsanity.com/paper/2503.23463/) (2025) - [Cosmos-Reason1: From Physical Common Sense To Embodied Reasoning](https://mlsanity.com/paper/2503.15558/) (2025) - [SimLingo: Vision-Only Closed-Loop Autonomous Driving with Language-Action Alignment](https://mlsanity.com/paper/2503.09594/) (2025) - [DriveAdapter: Breaking the Coupling Barrier of Perception and Planning in End-to-End Autonomous Driving](https://mlsanity.com/paper/2308.00398/) (2023) - [DriveTransformer: Unified Transformer for Scalable End-to-End Autonomous Driving](https://mlsanity.com/paper/2503.07656/) (2025) - [Are VLMs Ready for Autonomous Driving? An Empirical Study from the Reliability, Data, and Metric Perspectives](https://mlsanity.com/paper/2501.04003/) (2025) - [AlphaDrive: Unleashing the Power of VLMs in Autonomous Driving via Reinforcement Learning and Reasoning](https://mlsanity.com/paper/2503.07608/) (2025) - [Mini-InternVL: A Flexible-Transfer Pocket Multimodal Model with 5% Parameters and 90% Performance](https://mlsanity.com/paper/2410.16261/) (2024) - [OpenEMMA: Open-Source Multimodal Model for End-to-End Autonomous Driving](https://mlsanity.com/paper/2412.15208/) (2024) - [DriveCoT: Integrating Chain-of-Thought Reasoning with End-to-End Driving](https://mlsanity.com/paper/2403.16996/) (2024) - [Distilling Multi-modal Large Language Models for Autonomous Driving](https://mlsanity.com/paper/2501.09757/) (2025) - [DriveLMM-o1: A Step-by-Step Reasoning Dataset and Large Multimodal Model for Driving Scenario Understanding](https://mlsanity.com/paper/2503.10621/) (2025) - [HiP-AD: Hierarchical and Multi-Granularity Planning with Deformable Attention for Autonomous Driving in a Single Decoder](https://mlsanity.com/paper/2503.08612/) (2025) - [Hint-AD: Holistically Aligned Interpretability in End-to-End Autonomous Driving](https://mlsanity.com/paper/2409.06702/) (2024) - [Building Efficient Universal Classifiers with Natural Language Inference](https://mlsanity.com/paper/2312.17543/) (2023) - [VERDI: VLM-Embedded Reasoning for Autonomous Driving](https://mlsanity.com/paper/2505.15925/) (2025) - [Multi-Modal Adapter for Vision-Language Models](https://mlsanity.com/paper/2409.02958/) (2024) - [ALN-P3: Unified Language Alignment for Perception, Prediction, and Planning in Autonomous Driving](https://mlsanity.com/paper/2505.15158/) (2025) - [CogAD: Cognitive-Hierarchy Guided End-to-End Autonomous Driving](https://mlsanity.com/paper/2505.21581/) (2025) - [VLAD: A VLM-Augmented Autonomous Driving Framework with Hierarchical Planning and Interpretable Decision Process](https://mlsanity.com/paper/2507.01284/) (2025) ## Abstract (arXiv) Graph Visual Question Answering (GVQA) for autonomous driving organizes reasoning into ordered stages, namely Perception, Prediction, and Planning, where planning decisions should remain consistent with the model's own perception. We present a comparative study of cross-stage context passing on DriveLM-nuScenes using two complementary mechanisms. The explicit variant evaluates three prompt-based conditioning strategies on a domain-adapted 4B VLM (Mini-InternVL2-4B-DA-DriveLM) without additional training, reducing NLI contradiction by up to 42.6% and establishing a strong zero-training baseline. The implicit variant introduces gated context projectors, which extract a hidden-state vector from one stage and inject a normalized, gated projection into the next stage's input embeddings. These projectors are jointly trained with stage-specific QLoRA adapters on a general-purpose 8B VLM (InternVL3-8B-Instruct) while updating only approximately 0.5% of parameters. The implicit variant achieves a statistically significant 34% reduction in planning-stage NLI contradiction (bootstrap 95% CIs, p < 0.05) and increases cross-stage entailment by 50%, evaluated with a multilingual NLI classifier to account for mixed-language outputs. Planning language quality also improves (CIDEr +30.3%), but lexical overlap and structural consistency degrade due to the absence of driving-domain pretraining. Since the two variants use different base models, we present them as complementary case studies: explicit context passing provides a strong training-free baseline for surface consistency, while implicit gated projection delivers significant planning-stage semantic gains, suggesting domain adaptation as a plausible next ingredient for full-spectrum improvement.