- Rating
- 3(+1)
- Confidence
- 3
- Informativeness
- 3
- Presentation
- 3
- Soundness
- 1
Read review
Paper summary
DynaChain addresses the failure of fixed question-answering chains in closed-loop vision-language driving. A Qwen2.5-VL-3B agent selects relevant questions or commits to driving actions within a closed-loop decision process, and reinforcement learning combines route-level outcomes with dense step-wise feedback. On the ten-route Bench2Drive-VL evaluation, the method reports a Driving Score of 79.15 versus 58.93 without a reasoning chain and 42.24 with a fixed chain, using 2.52 questions per decision on average.
Critical strengths
- A meaningful failure mode motivates the work. Fixed reasoning chains may introduce irrelevant answers that distract a driving policy. Treating question selection as part of decision making addresses a practical limitation of vision-language driving agents.
- Closed-loop reasoning selection offers a coherent formulation. Modeling questions and action commitment as sequential decisions connects reasoning choices to later driving outcomes. This formulation moves beyond optimizing answer accuracy or reasoning length in isolation.
- Hybrid reinforcement learning provides a practical credit-assignment design. Combining route-level returns with step-wise control and infraction feedback gives useful signals at different time scales. Assigning delayed infraction penalties to preceding steps further targets decisions that contribute to failures.
- Reported driving improvements are substantial. Driving Score rises to 79.15, exceeding the no-chain baseline of 58.93 and reproduced AlphaDrive at 66.68. Average question usage also decreases to 2.52 from seven in the fixed-chain setting.
- Ablations examine several central modeling choices. Comparisons distinguish open-loop and closed-loop training, individual reward terms, and learned versus random question selection. Reported outcomes support the combined value of learned selection and closed-loop feedback.
Critical weaknesses
- Evaluation independence is unclear due to dev10-based model selection. Submitted training scripts describe using dev10 for validation and selecting checkpoints by its mean Driving Score, while the reported results also use dev10. Reuse of evaluation routes for model selection could overestimate generalization and requires an independent held-out test.
- A ten-route evaluation is insufficient for robust conclusions. Main driving metrics come from only ten scenarios, with a 50% success rate representing five successful routes. Additional routes, repeated seeds, and uncertainty estimates are needed to establish that the gains are stable.
- Privileged training answers introduce a train-test mismatch. Closed-loop reinforcement learning supplies expert answers to selected questions, whereas evaluation relies on the model’s own potentially inaccurate answers. Resulting selection policies may be vulnerable to answer noise or errors in safety-critical scenes.
- Benefits cannot be attributed exclusively to question selection. Reinforcement-learning updates also affect reasoning and committed driving actions rather than only the tool-selection policy. A frozen action-decision component would better isolate whether the reported gains come from choosing questions.
- More competitive low-cost selection baselines are needed. Existing controls cover random question subsets and selected adaptive reasoning systems but not simple risk-aware or uncertainty-triggered question selection. Such comparisons are important for establishing whether closed-loop RL is necessary to obtain similar decisions.
- End-to-end inference cost and deployment feasibility remain unclear. Each selected question requires an additional model call that processes the scene separately, despite the reduced question count. Without end-to-end latency and compute measurements, real-time driving practicality remains uncertain.
Questions for authors
- Did checkpoint selection or hyperparameter tuning use dev10? Please clarify the exact evaluation protocol and report results on an untouched route split.
- Can authors provide repeated-seed uncertainty and results on a larger, independent collection of driving scenarios?
- How does closed-loop training perform when question answers come from the model itself or are perturbed instead of supplied by an expert?
- Can the action-generation policy be frozen while training question selection to isolate the source of improvement?
- How do risk-aware or uncertainty-triggered selection baselines perform under a matched question budget?
- What are the complete per-decision latency and compute costs, including separate question-answering calls?
Thanks for the detailed response. The reported 4.41s per decision also remains far from real-time deployment. And comparing to Qwen model makes no sense. Although the method reduces latency relative to the fixed chain, the absolute cost still limits its practical value as a driving agent. I will keep my rating.