AutoVLA-D: Learn a Trajectory-Geometric Condition from Reasoning for Autonomous Driving VLA
Organizations: School of Electronic and Information Engineering, Beihang University · Institute for AI Industry Research (AIR), Tsinghua University · School of Computer Science and Engineering, Beihang University · National College for Excellent Engineers, Beihang University ·
Abstract
Autoregressive VLA planning usually predicts continuous trajectories as the final output, but driving instructions often describe actions rather than ego states, which weakens direct conditioning on the final continuous trajectory. We address this gap with TrajCond, which introduces a discrete trajectory-geometric condition for VLA planning. Specifically, we represent TrajCond as a 1D-Brohan differential action trajectory and learn it with a two-stage training framework before final continuous trajectory prediction. With the simultaneous autoregressive generation of discrete and numeric trajectories, we name our method AutoVLA-D (Double). Experiments on NAVSIM navtest, nuScenes, and NeuroNCAP show that discrete trajectories are weaker as direct outputs but more effective as TrajCond, consistently improving final continuous trajectory prediction across closed-loop and open-loop driving benchmarks. Code is available at https://github.com/ShivaTeam/AutoVLA-D.
Figures & tables
| Method | Base Model | NC | DAC | EP | TTC | C | PDMS |
| ImagiDrive-S | InternVL2.5-4B | 98.6 | 96.2 | 80.5 | 94.5 | 100.0 | 87.4 |
| ReCogDrive | InternVL3-8B | 98.2 | 97.8 | 83.5 | 95.2 | 100.0 | 89.6 |
| ReCogDrive † | Qwen3-VL-4B | 98.2 | 97.7 | 86.0 | 97.7 | 100.0 | 90.7 |
| DriveVLA-W0 | Emu3-8B | 98.7 | 99.1 | 83.3 | 95.3 | 99.3 | 90.2 |
| AutoVLA | Qwen2.5-VL-3B | 98.4 | 95.6 | 81.9 | 98.0 | 99.9 | 89.1 |
| AdaThinkDrive | InternVL3-8B | 98.4 | 97.8 | 84.4 | 95.2 | 100.0 | 90.3 |
| Method | Base Model | NC | DAC | DDC | TLC | EP | TTC | LK | C | EC | EPDMS |
| ReCogDrive | InternVL3-8B | 98.3 | 95.2 | 99.5 | 99.8 | 87.1 | 97.5 | 96.6 | 98.3 | 86.5 | 83.6 |
| DiffusionDrive | – | 98.2 | 95.9 | 99.4 | 99.8 | 87.5 | 97.3 | 96.8 | 98.3 | 87.7 | 84.5 |
| Curious-VLA | Qwen2.5-VL-3B | 98.4 | 96.9 | 99.2 | 99.8 | 88.5 | 97.9 | 96.9 | 98.1 | 81.5 | 85.3 |
| Curious-VLA † | Qwen3-VL-4B | 98.5 | 96.7 | 99.37 | 99.9 | 87.3 | 98.0 | 96.5 | 98.3 | 83.0 | 84.9 |
| DriveVLA-W0 | Emu3-8B | 98.5 | 99.1 | 98.0 | 99.7 | 86.4 | 98.1 | 93.2 | 97.9 | 58.9 | 86.1 |
| FLARE ∗ | Qwen3-VL-4B | 98.8 | 96.6 | 99.5 | 99.9 | 87.9 | 98.2 | 95.8 | 98.4 | 87.5 | 86.3 |
Appendix figures & tables12 assets
Supplementary material from the paper’s appendix.
Appendix
| Method | Base Model | NC | DAC | EP | TTC | C | PDMS |
| Ego-MLP | – | 93.1 | 78.3 | 63.2 | 84.0 | 100.0 | 66.4 |
| UniAD | – | 97.7 | 92.8 | 79.2 | 92.8 | 100.0 | 84.0 |
| DiffusionDrive | – | 98.2 | 96.2 | 82.2 | 94.7 | 100.0 | 88.1 |
| WoTE | – | 98.5 | 96.8 | 81.9 | 94.4 | 99.9 | 88.3 |
| Qwen2.5-VL | Qwen2.5-VL-7B | 97.8 | 92.1 | 78.3 | 92.8 | 100.0 | 83.3 |
| InternVL3 | InternVL3-8B | 97.0 | 92.4 | 78.9 | 91.8 | 100.0 | 83.3 |
| Method | NC | DAC | DDC | TLC | EP | TTC | LK | C | EC | EPDMS |
| Ego-MLP | 93.1 | 77.9 | 92.7 | 99.6 | 86.0 | 91.5 | 89.4 | 98.3 | 85.4 | 64.0 |
| VADv2 | 97.3 | 91.7 | 98.2 | 99.9 | 77.6 | 92.7 | 66.0 | 100.0 | 97.4 | 76.6 |
| TransFuser | 96.9 | 89.9 | 97.8 | 99.7 | 87.1 | 95.4 | 92.7 | 98.3 | 87.2 | 76.7 |
| HydraMDP++ | 97.2 | 97.5 | 99.4 | 99.6 | 83.1 | 96.5 | 94.4 | 98.2 | 70.9 | 81.4 |
| DriveSuprim | 97.5 | 96.5 | 99.4 | 99.6 | 88.4 | 96.6 | 95.5 | 98.3 | 77.0 | 83.1 |
| ARTEMIS | 98.3 | 95.1 | 98.6 | 99.8 | 81.5 | 97.4 | 96.5 | 98.3 | – | 83.1 |
| Method | Stage 1 | Stage 2 | RFT | EPDMS | PDMS |
| Brohan (B) | R B (6 ep.) | – | – | 82.4 | 86.8 |
| Continuous (C) | R C (6 ep.) | – | – | 83.9 | 88.5 |
| Brohan TrajCond (1) | – | B R C (6 ep.) | – | 84.0 | 89.2 |
| Brohan TrajCond | R B (3 ep.) | B R C (3 ep.) | – | 85.6 | 90.8 |
| Continuous + RFT | R C (6 ep.) | – | ✓ | 84.9 | 89.7 |
| Brohan TrajCond + RFT | R B (3 ep.) | B R C (3 ep.) | ✓ | 86.8 | 92.0 |
| Codebook | Tokenizer | Train error (cm) | Test error (cm) | ||||||
| 1s | 2s | 3s | 4s | 1s | 2s | 3s | 4s | ||
| 384 | 3D K-Means | 4.81 | 6.04 | 6.81 | 7.21 | 5.13 | 6.70 | 7.83 | 8.58 |
| 384 | 3D K-Disks | 6.37 | 8.71 | 10.03 | 10.77 | 7.06 | 10.20 | 12.29 | 13.49 |
| 384 | 1D-Brohan | 3.16 | 3.15 | 3.15 | 3.14 | 3.15 | 3.14 | 3.15 | 3.14 |
| 768 | 3D K-Means | 4.64 | 5.57 | 5.97 | 6.07 | 5.05 | 6.12 | 6.75 | 7.01 |
| 768 | 3D K-Disks | 4.25 | 5.11 | 5.53 | 5.78 | 4.69 | 5.84 | 6.65 | 7.09 |
| Codebook size | 3D K-Means | 3D K-Disks | 1D-Brohan |
| 384 | 0.0108958 | 0.0107432 | 0.0039045 |
| 768 | 0.0126990 | 0.0135494 | 0.0044001 |
| 1536 | 0.0167875 | 0.0175003 | 0.0051211 |
| Codebook | Representation | Direct output | TrajCond | ||
| EPDMS | PDMS | EPDMS | PDMS | ||
| 384 | 3D K-Means | 80.43 | 85.27 | – | – |
| 384 | 3D K-Disks | 81.44 | 86.61 | – | – |
| 384 | 1D-Brohan | 82.85 | 87.58 | – | – |
| 384 | Continuous | 83.90 | 88.49 | – | – |
| 768 | 3D K-Means | 80.69 | 85.78 | 85.65 | 90.72 |
| Setting | Condition | Target | Mean PPL |
| nuScenes | original | C | 1.0250 |
| nuScenes | B Pause | C | 1.0344 |
| NAVSIM | original | C | 1.0057 |
| NAVSIM | B Pause | C | 1.0068 |
| NAVSIM | R Pause | C | 1.0069 |
| NAVSIM | original | B | 1.0339 |
| Intent | NC zero count | DAC zero count | zero | ||||||||
| Ours | C | Reduction | Ours | C | Reduction | Ours | C | Ours rate | C rate | Reduction | |
| Go straight | 160 | 263 | 39.16% | 153 | 209 | 26.79% | 308 | 464 | 2.54% | 3.82% | 33.62% |
| Turn left | 29 | 52 | 44.23% | 151 | 200 | 24.50% | 178 | 246 | 1.47% | 2.03% | 27.64% |
| Turn right | 35 | 53 | 33.96% | 85 | 162 | 47.53% | 118 | 204 | 0.97% | 1.68% | 42.16% |
Autonomous Driving Planning
Autonomous driving planning selects safe, feasible vehicle actions and trajectories amid traffic, uncertainty, road rules, and changing goals.
See all 416 papers