Pith. sign in

REVIEW 4 major objections 8 minor 2 cited by

TeLL-Drive: Enhancing Autonomous Driving with Teacher LLM-Guided Deep Reinforcement Learning

T0 review · 4 major / 8 minor · reviewed 2026-08-09 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A teacher LLM guides a deep-reinforcement-learning student to success rates of 88–100% across three driving tasks.

desk verdict A sensible teacher-student idea with real success-rate gains, but the abstract overclaims on average returns and the KL constraint is under-specified. read the letter →

arxiv 2502.01387 v3 pith:KI6DFW63 submitted 2025-02-03 cs.AI cs.RO

Chengkai Xu , Jiaqi Liu , Shiyu Fang , Yiming Cui , Dong Chen , Peng Hang , Jian Sun This is my paper · ORCID
classification cs.AIcs.RO
keywords autonomousdrivinglargelanguagemodelsdeepreinforcementlearningteacher-studentself-attentionpolicyfusionsampleefficiencyvehicle-in-the-looptesting
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

TeLL-Drive claims that the two standard weaknesses of autonomous-driving agents—DRL's slow, sample-hungry learning and LLMs' slow, random decision-making—can be resolved by pairing them as teacher and student. A large language model is prompted with risk metrics, retrieved past scenarios, and driving heuristics to propose high-level maneuvers; an attention-based deep reinforcement learning agent uses those proposals as a soft constraint during early training and then acts as the fast final decision maker. The paper reports that this hybrid reaches 88% success in unsignalized left turns, 91% in high-speed ramp merging, and 100% in four-lane adaptive cruise, beating DRL baselines and the LLM-based Dilu baseline while keeping per-decision inference under about four milliseconds. If true, the framework shows that LLM common-sense reasoning can be used as a training-time bootstrap without sacrificing the real-time response that safety-critical driving demands.

What carries the argument

The load-bearing mechanism is the constrained fusion of two policies: a teacher policy $\pi_T$, approximated by a learned teacher embedding $h_t^T$ built from the LLM's discrete action suggestions, and a student policy $\pi_S$ built on its own embedding $h_t^S$. A multi-head self-attention layer computes $h_t = \alpha_t W_V h_t^T + h_t^S$ and feeds the fused representation into the student's policy and Q estimates, while Eq. (13) constrains the student by $D_{KL}(\pi_S(s_t), \pi_T(s_t)) \le \sigma$, with $\sigma$ small early and larger later. Around this core sit a risk-aware prompt constructor that uses TTCP, memory retrieval, and chain-of-thought reasoning; a reflective evaluator that flags episodes with $\max_t \Omega(s_t, a_t) \ge \delta$ and writes revised policies back to memory; and an auxiliary distillation loss that teaches the teacher embedding to reproduce the LLM's choices. The attention mechanism is what lets the student keep its own learned strategy instead of being overwritten by the teacher.

What would settle it

Hold everything fixed except replace the LLM teacher's advice with uniformly random maneuvers drawn from the same five-action set, and retrain the student under the same KL schedule and attention architecture. If the random-teacher version still reaches roughly 88%, 91%, and 100% success in the three scenarios, then the content of the LLM's reasoning is not what produces the reported gains; if the random-teacher version collapses, the teacher's semantic guidance is load-bearing.

Watch

Extended reading notes

Core claim

The paper's central claim is that high-level LLM reasoning and low-latency DRL control are complements, not competitors, and that a teacher-student architecture can capture both. The LLM teacher never drives the vehicle; it produces contextual maneuver proposals from prompts built on a Time-to-Conflict-Point risk metric, cosine-similarity retrieval from a memory repository, and chain-of-thought reasoning. The DRL student is an actor-critic whose objective includes a KL-divergence constraint (Eq. 13) that keeps the student policy near the teacher's proposals early in training, with the bound loosened as learning proceeds; a multi-head self-attention layer fuses a learned teacher embedding with the student's own state embedding, and an auxiliary distillation loss (Eq. 22) makes the teacher embedding approximate the LLM's implicit policy. The paper reports that the student surpasses the teacher in all three test scenarios and that the full framework beats DQN, A2C, RecurrentPPO, and Dilu on success rate and return while keeping decision time at 0.003–0.004 seconds, and that vehicle-in-loop experiments at an unsignalized intersection confirm safe yielding and priority-taking behaviors on real hardware.

Load-bearing premise

The load-bearing premise is that the LLM's discrete action outputs can be converted into a well-defined teacher distribution $\pi_T$ for the KL constraint in Eq. (13), but the paper never specifies how that distribution is formed.

Editorial extensions

If this is right

  • If the reported results hold, autonomous driving planners can get LLM-level reasoning without LLM latency: the teacher is used only in the first 10% of training, and the student makes decisions in about 0.004 seconds.
  • Training data requirements shrink because the KL constraint steers exploration early, which addresses DRL's sample-complexity bottleneck in safety-critical scenarios.
  • The teacher-student split means the expensive LLM can be swapped or replaced after training, and the deployed vehicle needs only the student policy network.
  • The student's consistent outperformance of the teacher implies LLM guidance is a bootstrap rather than a performance ceiling, so the framework can in principle improve as the DRL agent accumulates experience.
  • Vehicle-in-loop demonstrations at an unsignalized intersection suggest the approach transfers from simulation to real hardware for at least one high-risk scenario.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A testable extension the paper leaves implicit: any cheap expert signal—rule-based heuristics, human demonstrations, or a smaller model—might yield most of the same early-training boost, since the attention and KL mechanism do not logically require an LLM; comparing teacher sources would isolate what the LLM's reasoning adds.
  • The paper's KL constraint requires a distribution $\pi_T$ that the LLM never directly emits; until the paper specifies whether $\pi_T$ is a one-hot over the chosen action, a token-probability softmax, or a separately trained estimator, the mechanism in Eq. (13) is underdetermined and reimplementation is not uniquely defined.
  • The memory repository and reflective evaluator produce a trace of which past scenarios and risk thresholds influenced each maneuver, which could be repurposed as an interpretability and safety-audit log for the deployed policy.
  • Because the student surpasses the teacher, the framework suggests a general pattern: expert guidance should be scheduled to decay, and the final policy should be evaluated independently of the teacher; future multi-agent or open-road studies could test whether this decay schedule needs retuning in more interactive traffic.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 8 minor

Summary. The paper proposes TeLL-Drive, a hybrid decision-making framework for autonomous driving in which a GPT-4o-mini-based Teacher LLM provides high-level action guidance to an attention-based student deep reinforcement learning (DRL) agent. The method constructs risk-aware prompts using time-to-conflict-point estimates, memory retrieval, and chain-of-thought reasoning, and constrains the student policy toward the teacher's demonstrations via a KL-divergence term, with a learned approximation of the teacher's policy. The authors evaluate the method in three Highway-Env scenarios (unsignalized intersection, ramp merging, four-lane adaptive cruise) against DQN, A2C, RecurrentPPO, and the LLM-based Dilu, and report simulation results, an ablation study, and a qualitative vehicle-in-loop experiment. The main claimed contributions are improved success rate, average return, and real-time feasibility over existing baselines.

Significance. If the performance claims were fully supported, the idea of using an LLM teacher to accelerate DRL convergence and improve robustness in driving scenarios would be of practical interest to the autonomous-driving and robot-learning communities. The paper includes a useful comparison against both conventional DRL and an LLM-based baseline, a clear three-scenario evaluation, and an ablation that isolates the contribution of the attention mechanism and the LLM guidance. The vehicle-in-loop setup is a positive step toward real-world validation. However, the current manuscript does not substantiate the headline claim on average returns, and the undefined teacher distribution and reward function prevent the method from being reproduced or properly assessed. With these gaps closed, the framework could become a solid contribution; in its present form the evidence is insufficient.

major comments (4)
  1. [Section V.C.1, Table II] The abstract claims that TeLL-Drive outperforms baselines "in terms of success rates, average returns, and real-time feasibility," but Table II shows Dilu attaining higher Eval Reward in both the unsignalized intersection (6.59 vs. 5.68) and four-lane adaptive cruise (29.53 vs. 27.17) scenarios. The text introducing Table II calls these columns "Evaluation Return," which is the paper's apparent measure of average return. The authors must clarify what "average returns" in the abstract refers to, and either correct the claim, re-run the comparison, or present a consistent metric across all scenarios that supports the stated superiority.
  2. [Section IV.C, Eq. (13)] The constrained optimization uses a KL divergence between the student policy πS(s_t) and a teacher policy πT(s_t), but Algorithm 1 outputs only a discrete action a_t ∈ A, not a probability distribution over the action space. Section IV.C.2 explicitly states that the teacher does not directly provide action probabilities and instead introduces a learned approximation π̂T via Eq. (15). The paper never specifies how πT in Eq. (13) is obtained (e.g., one-hot distribution, token-level softmax, or the learned π̂T), nor how the constraint is enforced during optimization. Without this, the central guidance mechanism is not implementable as described.
  3. [Section III and Section IV] The reward function r(s_t, a_t) is listed as part of the POMDP tuple in Section III but is never defined or described anywhere in the manuscript. This leaves the "Eval Reward" in Table II and the "average return" in Fig. 5 without a verifiable meaning, and prevents the reader from confirming that all baselines were optimized under the same reward. The authors should provide the reward specification (or explicitly reference the Highway-Env default reward and any modifications) and state whether the same reward is used for the teacher's guidance and the student's objective.
  4. [Section IV.B.1, Eq. (4) and Algorithm 1] The TTCP (time-to-conflict-point) metric is defined inconsistently. Eq. (4) writes τ = arg min_{t≥0} || p_ego(t)/v_ego(t) − p_other(t)/v_other(t) ||, which is dimensionally ambiguous and not a standard collision measure, while Algorithm 1 line 4 uses τ_i = arg min_{Δt≥0} || p_ego(t+Δt) − p_i(t+Δt) ||, a different quantity. Since TTCP drives the risk metric in Eq. (8) and the teacher's prompting, this inconsistency must be corrected and a precise, implementable definition provided.
minor comments (8)
  1. [Section I, Section V.A] The phrase "gradient verification scenario" in the Introduction and Section V.A appears to be a typo; it should probably read "graded verification scenario" or "progressively difficult scenario."
  2. [Table I vs. Eqs. (8) and (13)] Table I does not report values for the risk weighting β, the high-risk threshold δ, or the KL tolerance σ, even though these parameters appear in the algorithm's key equations; please include them for reproducibility.
  3. [Eq. (13) and surrounding text] The notation for the teacher policy is inconsistent: Eq. (13) and its caption use πT, while the immediately following sentence refers to πE; standardize the symbol.
  4. [Abstract and Section VI] The Abstract and Section VI refer to "Our Website" and a footnote reads "Vehicle-in-Loop Experimental Validation Video Weblink," but no actual URL is provided; include a working link or remove the reference.
  5. [Fig. 9 caption] The caption contains a typo: "TeLL-Dirve" should be "TeLL-Drive."
  6. [Section IV.B.3] The text says "proposes a domin-specific adjustment"; this should be "domain-specific adjustment."
  7. [Section V.C.2] The ablation study is presented only as training curves in Fig. 6 without a quantitative table of final rewards or success rates; adding end-of-training numbers would strengthen the comparison.
  8. [Section VI.B] The vehicle-in-loop experiments are reported qualitatively through two cases; no quantitative metrics (e.g., success rate, task completion time, safety margin) are provided for the real-vehicle runs, so the claim of "robustness and reliability" is not quantified.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the teacher LLM is an external model, the student objective is compositionally defined, and the reported metrics are external benchmark measurements rather than fitted inputs relabeled as predictions.

full rationale

The paper's derivation chain is self-contained in the relevant sense. The teacher guidance comes from GPT-4o-mini, an external LLM whose parameters are not fitted to the success rates or returns that are later reported; the student is a standard actor-critic augmented with a self-attention fusion mechanism. The KL-constrained objective in Eq. (13) and the distillation loss in Eq. (22) define a training procedure rather than a post-hoc reproduction of the evaluation numbers. The headline quantities in Table II — success rate, Eval Reward, average speed, Delta-TTCP, and consumption time — are measured against external baselines in Highway-Env and in a vehicle-in-loop experiment, so they do not reduce by construction to any fitted constant. The paper does contain self-citations, notably [23] for the TTCP risk metric and [27] as related work on LLM-guided distillation, but these are attributional and contextual: the TTCP definition is restated in Eq. (4), and [27] is not invoked as a uniqueness theorem or as the sole justification for the method. The underspecification of the teacher distribution pi_T in Eq. (13) and the absence of a stated reward function are implementation gaps, not circularity, because an undefined object cannot make a later result equivalent to an earlier input. The apparent mismatch between the abstract's 'average returns' claim and Dilu's higher Eval Reward in Table II is a correctness or reporting risk, not a circularity, since the table is an external measurement rather than an assumed conclusion. No fitted parameter is renamed as a prediction, and no load-bearing premise is justified solely by a self-citation chain. Accordingly, no circular steps are identified.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The framework uses an external LLM (GPT-4o-mini) and standard DRL components; no new physical entities are introduced. The key unstated inputs are the reward function, the teacher policy distribution, and several hyperparameters (β, δ, σ), which are necessary to reproduce the method.

free parameters (4)
  • β (risk weighting in Eq. 8)
    Weighting constant for the infraction indicator in the risk function; value not specified, affects when episodes are flagged for reflection.
  • δ (high-risk threshold in Eq. 8 / Algorithm 1)
    Threshold for flagging segments for reflection; value not given, controls how often reflective feedback is applied.
  • σ (KL tolerance in Eq. 13)
    Initial value and schedule for the KL bound not specified; directly controls how strongly teacher guidance constrains the student policy.
  • Prompt template and CoT instructions
    Free-form design choices for the LLM prompts that determine the teacher's behavior; not disclosed, so the teacher's actions are not reproducible.
assumptions (4)
  • domain assumption Highway-Env with three scenarios is a representative testbed for autonomous driving decision-making
    The paper evaluates only in this simulator; no real-world quantitative validation is provided (Section V-A).
  • domain assumption The LLM (GPT-4o-mini) provides reliable, consistent chain-of-thought guidance
    The method depends on the teacher's outputs being sensible; the paper does not measure LLM output reliability or variance (Section IV-B).
  • domain assumption The TTCP risk metric accurately captures collision risk
    TTCP is used in prompts and risk reflection, but Eq. (4) is dimensionally inconsistent with Algorithm 1 line 4 (Section IV-B).
  • ad hoc to paper The teacher policy πT can be represented as a distribution for the KL constraint
    No procedure is given to convert discrete LLM actions into a policy distribution; Eq. (13) requires πT (Section IV-C1).

how reviews work

0 comments
Cite this review

Pith. "Pith review of TeLL-Drive: Enhancing Autonomous Driving with Teacher LLM-Guided Deep Reinforcement Learning." pith.science (2026). https://pith.science/paper/KI6DFW63

@misc{pith2026250201387,
  author       = {Pith},
  title        = {Pith review of: TeLL-Drive: Enhancing Autonomous Driving with Teacher LLM-Guided Deep Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KI6DFW63}},
  note         = {Machine review of arXiv:2502.01387}
}
read the original abstract

Although Deep Reinforcement Learning (DRL) and Large Language Models (LLMs) each show promise in addressing decision-making challenges in autonomous driving, DRL often suffers from high sample complexity, while LLMs have difficulty ensuring real-time decision making. To address these limitations, we propose TeLL-Drive, a hybrid framework that integrates a Teacher LLM to guide an attention-based Student DRL policy. By incorporating risk metrics, historical scenario retrieval, and domain heuristics into context-rich prompts, the LLM produces high-level driving strategies through chain-of-thought reasoning. A self-attention mechanism then fuses these strategies with the DRL agent's exploration, accelerating policy convergence and boosting robustness across diverse driving conditions. The experimental results, evaluated across multiple traffic scenarios, show that TeLL-Drive outperforms existing baseline methods, including other LLM-based approaches, in terms of success rates, average returns, and real-time feasibility. Ablation studies underscore the importance of each model component, especially the synergy between the attention mechanism and LLM-driven guidance. Finally, we build a virtual-real fusion experimental platform to verify the real-time performance, robustness, and reliability of the algorithm running on real vehicles through vehicle-in-loop experiments.

Figures

Figures reproduced from arXiv: 2502.01387 by the authors.

Figure 1
Figure 1. The LLM teacher guides the DRL agent in decision-making within [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall conceptual framework of TeLL-Drive, where a DRL student agent is guided by the LLM teacher for better decision making in autonomous [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Proposed policy network with self-attention layer. The network [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: The designed gradient verification scenario for simulation: (a) [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Comparison of the performance of this model with traditional DRL training results. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Comparison of performance results during the ablation experiment training process. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Comparison of testing success rate results between the teacher agent [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Test case performance results of TeLL-Drive in three scenarios (a) Unsignalized Intersection, (b) High-Speed Ramp Merging, (c) Four-Lane Adaptive [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: The virtual-reality fusion experimental platform built based on Tongji [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: A real vehicle-in-loop experiment based on the virtual-reality fusion experimental platform. The vehicle equipped with TeLL-Drive choose to yield [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. A Survey on the Applications of Generative Artificial Intelligence in Automated Driving Systems Test Scenario Generation Methods

    cs.SE 2025-12 reject novelty 4.0 of 10

    A literature survey of scenario-generation methods for ADS testing that adds an unvalidated AII/RAS/OCS metric suite and ODD-difficulty schema, undermined by inconsistent calculations in the worked examples.

  2. LeAD: The LLM Enhanced Planning System Converged with End-to-end Autonomous Driving

    cs.RO 2025-07 conditional novelty 4.0 of 10

    LeAD adds a low-frequency large-language-model planner that takes over when a high-frequency end-to-end driving model gets stuck, and reports improved CARLA benchmark scores.

Reference graph

Works this paper leans on

37 extracted references · 19 canonical work pages · cited by 2 Pith papers

  1. [1]

    Milestones in autonomous driving and intelligent vehicles: Survey of surveys

    Long Chen, Yuchen Li, Chao Huang, Bai Li, Yang Xing, Daxin Tian, Li Li, Zhongxu Hu, Xiaoxiang Na, Zixuan Li, et al. Milestones in autonomous driving and intelligent vehicles: Survey of surveys. IEEE Transactions on Intelligent Vehicles, 8(2):1046–1056, 2022

    2022

  2. [2]

    A survey of end-to-end driving: Architectures and training methods

    Ardi Tampuu, Tambet Matiisen, Maksym Semikin, Dmytro Fishman, and Naveed Muhammad. A survey of end-to-end driving: Architectures and training methods. IEEE Transactions on Neural Networks and Learning Systems, 33(4):1364–1384, 2020

    work page 2020

  3. [3]

    Survey of deep reinforcement learning for motion planning of autonomous vehicles

    Szil ´ard Aradi. Survey of deep reinforcement learning for motion planning of autonomous vehicles. IEEE Transactions on Intelligent Transportation Systems, 23(2):740–759, 2020

    work page 2020

  4. [4]

    Deep reinforcement learning for intelligent transportation systems: A survey

    Ammar Haydari and Yasin Yılmaz. Deep reinforcement learning for intelligent transportation systems: A survey. IEEE Transactions on Intelligent Transportation Systems, 23(1):11–32, 2020

    work page 2020

  5. [5]

    Mtd- gpt: A multi-task decision-making gpt model for autonomous driving at unsignalized intersections

    Jiaqi Liu, Peng Hang, Xiao Qi, Jianqiang Wang, and Jian Sun. Mtd- gpt: A multi-task decision-making gpt model for autonomous driving at unsignalized intersections. In 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC) , pages 5154–5161. IEEE, 2023

    work page 2023

  6. [6]

    Decision making of autonomous vehicles in lane change scenarios: Deep reinforcement learning approaches with risk awareness

    Guofa Li, Yifan Yang, Shen Li, Xingda Qu, Nengchao Lyu, and Shengbo Eben Li. Decision making of autonomous vehicles in lane change scenarios: Deep reinforcement learning approaches with risk awareness. Transportation research part C: emerging technologies , 134:103452, 2022

    work page 2022

  7. [7]

    Automatically generated curriculum based reinforcement learning for autonomous vehicles in urban environment

    Zhiqian Qiao, Katharina Muelling, John M Dolan, Praveen Palanisamy, and Priyantha Mudalige. Automatically generated curriculum based reinforcement learning for autonomous vehicles in urban environment. In 2018 IEEE Intelligent Vehicles Symposium (IV) , pages 1233–1238. IEEE, 2018

    work page 2018

  8. [8]

    Cooperation-aware reinforcement learning for merging in dense traffic

    Maxime Bouton, Alireza Nakhaei, Kikuo Fujimura, and Mykel J Kochenderfer. Cooperation-aware reinforcement learning for merging in dense traffic. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC), pages 3441–3447. IEEE, 2019

    2019

Show all 37 references
  1. [9]

    Formulation of deep reinforcement learning architecture toward autonomous driving for on-ramp merge

    Pin Wang and Ching-Yao Chan. Formulation of deep reinforcement learning architecture toward autonomous driving for on-ramp merge. In 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC), pages 1–6. IEEE, 2017

    2017

  2. [10]

    Deep reinforcement learning for autonomous driving: A survey

    B Ravi Kiran, Ibrahim Sobh, Victor Talpaert, Patrick Mannion, Ahmad A Al Sallab, Senthil Yogamani, and Patrick P ´erez. Deep reinforcement learning for autonomous driving: A survey. IEEE Transactions on Intelligent Transportation Systems, 23(6):4909–4926, 2021

    2021

  3. [11]

    Gpt-4o system card

    Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, et al. Gpt-4o system card. arXiv preprint arXiv:2410.21276 , 2024

    2024 arXiv

  4. [12]

    Llm4drive: A survey of large language models for autonomous driving

    Zhenjie Yang, Xiaosong Jia, Hongyang Li, and Junchi Yan. Llm4drive: A survey of large language models for autonomous driving. In NeurIPS 2024 Workshop on Open-World Agents , 2023

    2024

  5. [13]

    A survey on multimodal large language models for autonomous driving

    Can Cui, Yunsheng Ma, Xu Cao, Wenqian Ye, Yang Zhou, Kaizhao Liang, Jintai Chen, Juanwu Lu, Zichong Yang, Kuei-Da Liao, et al. A survey on multimodal large language models for autonomous driving. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Visi...

    2024

  6. [14]

    Drivegpt4: Inter- pretable end-to-end autonomous driving via large language model

    Zhenhua Xu, Yujia Zhang, Enze Xie, Zhen Zhao, Yong Guo, Kwan- Yee K Wong, Zhenguo Li, and Hengshuang Zhao. Drivegpt4: Inter- pretable end-to-end autonomous driving via large language model. IEEE Robotics and Automation Letters , 2024

    2024

  7. [15]

    Drivellm: Charting the path toward full autonomous driving with large language models

    Yaodong Cui, Shucheng Huang, Jiaming Zhong, Zhenan Liu, Yutong Wang, Chen Sun, Bai Li, Xiao Wang, and Amir Khajepour. Drivellm: Charting the path toward full autonomous driving with large language models. IEEE Transactions on Intelligent Vehicles , 2023

    2023

  8. [16]

    Cooperative decision-making for cavs at unsignalized intersections: A marl approach with attention and hierarchical game priors

    Jiaqi Liu, Peng Hang, Xiaoxiang Na, Chao Huang, and Jian Sun. Cooperative decision-making for cavs at unsignalized intersections: A marl approach with attention and hierarchical game priors. IEEE Transactions on Intelligent Transportation Systems , 2024

    2024

  9. [17]

    A reinforcement learning approach to autonomous decision making of intelligent vehicles on highways

    Xin Xu, Lei Zuo, Xin Li, Lilin Qian, Junkai Ren, and Zhenping Sun. A reinforcement learning approach to autonomous decision making of intelligent vehicles on highways. IEEE Transactions on Systems, Man, and Cybernetics: Systems , 50(10):3884–3897, 2018

    2018

  10. [18]

    Proximal policy optimization algorithms

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347, 2017

    2017 arXiv

  11. [19]

    Playing atari with deep reinforcement learning

    V olodymyr Mnih. Playing atari with deep reinforcement learning. arXiv preprint arXiv:1312.5602, 2013

    2013 arXiv

  12. [20]

    A survey of deep rl and il for autonomous driving policy learning

    Zeyu Zhu and Huijing Zhao. A survey of deep rl and il for autonomous driving policy learning. IEEE Transactions on Intelligent Transportation Systems, 23(9):14043–14065, 2021

    2021

  13. [21]

    Aligning large multimodal models with factually augmented rlhf

    Zhiqing Sun, Sheng Shen, Shengcao Cao, Haotian Liu, Chunyuan Li, Yikang Shen, Chuang Gan, Liang-Yan Gui, Yu-Xiong Wang, Yiming Yang, et al. Aligning large multimodal models with factually augmented rlhf. arXiv preprint arXiv:2309.14525 , 2023

    2023 arXiv

  14. [22]

    Rlhf-v: Towards trustworthy mllms via behavior alignment from fine- grained correctional human feedback

    Tianyu Yu, Yuan Yao, Haoye Zhang, Taiwen He, Yifeng Han, Ganqu Cui, Jinyi Hu, Zhiyuan Liu, Hai-Tao Zheng, Maosong Sun, et al. Rlhf-v: Towards trustworthy mllms via behavior alignment from fine- grained correctional human feedback. In Proceedings of the IEEE/CVF Conference on C...

    2024

  15. [23]

    Towards interactive and learnable cooperative driving automation: a large language model-driven decision-making framework

    Shiyu Fang, Jiaqi Liu, Mingyu Ding, Yiming Cui, Chen Lv, Peng Hang, and Jian Sun. Towards interactive and learnable cooperative driving automation: a large language model-driven decision-making framework. arXiv preprint arXiv:2409.12812 , 2024

    2024

  16. [24]

    Languagempc: Large language models as decision makers for au- tonomous driving

    Hao Sha, Yao Mu, Yuxuan Jiang, Li Chen, Chenfeng Xu, Ping Luo, Shengbo Eben Li, Masayoshi Tomizuka, Wei Zhan, and Mingyu Ding. Languagempc: Large language models as decision makers for au- tonomous driving. arXiv preprint arXiv:2310.03026 , 2023

    arXiv

  17. [25]

    Drive like a human: Rethinking autonomous driving with large language models

    Daocheng Fu, Xin Li, Licheng Wen, Min Dou, Pinlong Cai, Botian Shi, and Yu Qiao. Drive like a human: Rethinking autonomous driving with large language models. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , pages 910–919, 2024

    2024

  18. [26]

    Dilu: A knowledge- driven approach to autonomous driving with large language models

    Licheng Wen, Daocheng Fu, Xin Li, Xinyu Cai, Tao Ma, Pinlong Cai, Min Dou, Botian Shi, Liang He, and Yu Qiao. Dilu: A knowledge- driven approach to autonomous driving with large language models. arXiv preprint arXiv:2309.16292 , 2023

    2023 arXiv

  19. [27]

    Language-driven policy distillation for cooperative driving in multi-agent reinforcement learning

    Jiaqi Liu, Chengkai Xu, Peng Hang, Jian Sun, Mingyu Ding, Wei Zhan, and Masayoshi Tomizuka. Language-driven policy distillation for cooperative driving in multi-agent reinforcement learning. arXiv preprint arXiv:2410.24152, 2024

    arXiv

  20. [28]

    Large language models are semi-parametric reinforcement learning agents

    Danyang Zhang, Lu Chen, Situo Zhang, Hongshen Xu, Zihan Zhao, and Kai Yu. Large language models are semi-parametric reinforcement learning agents. Advances in Neural Information Processing Systems , 36, 2024

    2024

  21. [29]

    Automl-agent: A multi-agent llm framework for full-pipeline automl

    Patara Trirat, Wonyong Jeong, and Sung Ju Hwang. Automl-agent: A multi-agent llm framework for full-pipeline automl. arXiv preprint arXiv:2410.02958, 2024

    2024 arXiv

  22. [30]

    Eureka: Human-level reward design via coding large language models

    Yecheng Jason Ma, William Liang, Guanzhi Wang, De-An Huang, Osbert Bastani, Dinesh Jayaraman, Yuke Zhu, Linxi Fan, and Anima Anandkumar. Eureka: Human-level reward design via coding large language models. arXiv preprint arXiv:2310.12931 , 2023

    arXiv

  23. [31]

    Chain-of-thought prompting elicits reasoning in large language models

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural infor- mation processing systems , 35:24824–24837, 2022

    2022

  24. [32]

    A survey of actor-critic reinforcement learning: Standard and natural policy gradients

    Ivo Grondman, Lucian Busoniu, Gabriel AD Lopes, and Robert Babuska. A survey of actor-critic reinforcement learning: Standard and natural policy gradients. IEEE Transactions on Systems, Man, and Cybernetics, part C (applications and reviews) , 42(6):1291–1307, 2012

    2012

  25. [33]

    Efficient deep reinforcement learning with imitative expert priors for autonomous driving

    Zhiyu Huang, Jingda Wu, and Chen Lv. Efficient deep reinforcement learning with imitative expert priors for autonomous driving. IEEE Transactions on Neural Networks and Learning Systems , 34(10):7391– 7403, 2022

    2022

  26. [34]

    An environment for autonomous driving decision- making

    Edouard Leurent. An environment for autonomous driving decision- making. https://github.com/eleurent/highway-env, 2018

    2018

  27. [35]

    Asynchronous methods for deep reinforcement learning

    V olodymyr Mnih. Asynchronous methods for deep reinforcement learning. arXiv preprint arXiv:1602.01783 , 2016

    2016 arXiv

  28. [36]

    Generalization, mayhems and limits in recurrent proximal policy op- timization

    Marco Pleines, Matthias Pallasch, Frank Zimmer, and Mike Preuss. Generalization, mayhems and limits in recurrent proximal policy op- timization. arXiv preprint arXiv:2205.11104 , 2022

    2022 arXiv

  29. [37]

    Modeling and simulation of merging behavior at urban expressway on-ramp

    Zuo Kang, Qiyuan Liu, and Sun Jian. Modeling and simulation of merging behavior at urban expressway on-ramp. Journal of System Simulation, 29(9):1895–1906, 2020

    1906

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.