Pith. sign in

REVIEW 4 major objections 3 minor 59 references

Query-Efficient Planning with Language Models

T0 review · 4 major / 3 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read Generative LLM planners that propose full plans and replan from world-model feedback beat LLM-guided heuristic search, reaching 78-89% success under a 20-query budget.

desk verdict A useful empirical comparison with a real confound: Boomerang gets full interaction history while ToI does not, so the headline adaptivity claim is only as strong as that asymmetry. read the letter →

arxiv 2412.06162 v1 pith:JROXP5PK submitted 2024-12-09 cs.AI cs.CL

classification cs.AIcs.CL
keywords query-efficientplanninglargelanguagemodelsgenerativeLLMasheuristicworld-modelquerieslazysearchreplanningwithfeedbackPlanBench
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

When a planner's dominant cost is checking transitions against a world model, the paper claims, how the LLM is attached matters more than raw LLM reasoning power. It proposes two competing designs: Tree of Interaction, which uses the LLM as a heuristic that proposes actions and ranks states inside a BFS/DFS search, and Boomerang, which uses the LLM as a generative planner that emits a full start-to-goal action sequence, verifies it against the world model, and replans from the error feedback. The paper's central finding is that the generative design is more query-efficient: it reaches 78% success on 600 Blocksworld problems, 82% on Logistics, 89% on Grippers, and 57% on Robotouille within a 20-query budget, versus 63% for the best classical planner and 37% for the best LLM-heuristic variant on Blocksworld. The explanation offered is adaptivity: a generative planner resets to the start and can route around a dead end in its next whole plan, while a heuristic can only pick among nodes the search tree happens to offer. The result matters because in robotics and task-and-motion planning, world-model queries are often the expensive step, so a planner that reaches the goal in fewer queries could translate directly into faster real-time planning.

What carries the argument

The central mechanism is the Boomerang loop: generate a complete action sequence with the LLM, verify it by querying the world model on every transition, feed back the partial trajectory and error, then regenerate, with all previously verified transitions cached so each iteration costs only the newly revealed states. This is the paper's instantiation of lazy search, where the planner reasons over a cheap internal model and only pays for edges on promising paths; the paper formalizes the LLM as a posterior sampler over world models $P_\theta(M^\star \mid \phi, H_t)$ and adapts a Bayesian regret bound of $O(\tau\sqrt{SAT\log(SAT)})$ on the infeasible edges queried before a feasible path is found. The contrasting object is ToI, which embeds the LLM as an action proposer and a three-class state evaluator inside beam search or DFS, so the LLM's influence is filtered through the search tree and its state rankings, which the paper notes are inadmissible and inconsistent because states are ranked independently and history is discarded each iteration.

What would settle it

Re-run the same planners on a domain where each world-model query is genuinely expensive, for example collision-checking a robot arm or running a local motion optimizer, charge the budget for every transition accessed from cache, and compare wall-clock time or paid queries instead of unique transition counts; if Boomerang's margin over Classical and ToI shrinks or reverses under that accounting, the query-efficiency claim rests on the free-cache assumption rather than on generative adaptivity.

Watch

Extended reading notes

Core claim

The paper's central claim, stated in its abstract, is that using an LLM as a generative planner results in significantly fewer world-model interactions than using an LLM as a heuristic, and that the generative planner can more rapidly adapt its planning strategies based on immediate feedback. Boomerang operationalizes this by prompting the LLM to produce an entire plan from the initial state to the goal, rolling that plan out through a deterministic world model, and appending the verified prefix plus the error message to the LLM's context before generating the next plan; ToI instead confines the LLM to proposing actions and rating states as impossible/maybe/sure within a beam or depth-first search. Under a 20-world-model-query budget on 600 PlanBench Blocksworld problems and 100 problems each in Logistics, Grippers, and Robotouille, Boomerang solves them at rates of 78%, 82%, 89%, and 57%, respectively, while the strongest classical planner configuration reaches 63% on Blocksworld and 5-13% on the other PDDL domains, and the best ToI variant reaches 37% on Blocksworld. The paper also reports that Boomerang uses fewer LLM calls (5.69 per problem versus 13.69 for ReAct and 28-40 for the ToI variants) and achieves an optimality rate of 0.69, second only to the optimality-guaranteed classical planner at 0.63. The authors connect this performance to lazy search and posterior sampling, adapting a Bayesian regret bound of $O(\tau\sqrt{SAT\log(SAT)})$ on the number of infeasible transitions queried before a feasible plan is found.

Load-bearing premise

The comparison prices a world-model query as the scarce resource and lets Boomerang reuse every previously verified transition for free, so the finding holds only if re-checking a cached transition costs nothing and if world-model queries, not LLM API calls or other computation, are the true bottleneck; the paper asserts but does not demonstrate this on an actually expensive world model.

Editorial extensions

If this is right

  • In planning domains where world-model queries dominate cost, the generative-planner interface should be the default: Boomerang needs only 5.69 LLM calls per Blocksworld problem versus 13.69 for ReAct and 28-40 for ToI variants, and it reaches the goal in fewer queries on average (12.15) than both ReAct (13.06) and the best classical planner (14.67).
  • The gap should grow with a domain's branching factor: the paper's lazy-search analysis shows that even a perfect heuristic run as A* pays $k$ times more edge queries than a lazy planner over the optimal path, which is consistent with Grippers, the domain with the largest action space, showing Boomerang's widest margin (89% versus 31% for ToI-DFS and 13% for Classical).
  • The adaptivity explanation predicts that any method that resets to the start but still acts one step at a time will underperform whole-plan regeneration, which is exactly the appendix finding that Reflexion outperforms ReAct but not Boomerang.
  • Success within a fixed query budget and solution optimality are partly independent: the I/O + CoT + P ablation, which repeats the goal after every step, beats ReAct by 14.8% without any interaction loop, showing that prompt design and interaction design are separate levers.

Reading between the lines

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

  • If the adaptivity story is right, then compressing the LLM's history, replacing raw trajectory dumps with a distilled summary of what failed and why, should extend Boomerang's advantage to long-horizon problems, where the paper's own limitations section concedes it begins reproposing failed sequences through context forgetting.
  • The free-cache accounting is testable: a deployment that charges for every world-model access, including re-verifying a cached transition, or one whose bottleneck is LLM API latency rather than world-model queries, could erase or reverse Boomerang's margin, since the paper asserts rather than measures the claimed correlation between query count and wall-clock time.
  • A natural hybrid suggested by the comparison is to seed a classical beam search with Boomerang's generated plans, giving the tree-based framework the cul-de-sac escape it currently lacks while preserving search structure; the paper does not explore this direction.
  • Re-running the comparison under a budget of plan length or total LLM tokens rather than verified transitions would separate the value of whole-plan generation from the value of start-state resets, which the paper treats as a single design.
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 / 3 minor

Summary. The paper proposes two ways of using LLMs for planning under a limited budget of world-model queries: Tree of Interaction (ToI), where an LLM acts as an action proposer and state evaluator inside BFS/DFS search, and Boomerang, where an LLM generates complete action sequences and replans from world-model feedback. Experiments on 600 PlanBench Blocksworld problems and 100-problem Logistics, Grippers, and Robotouille suites report that Boomerang attains the highest success rates under a 20-query cap, with 78% on Blocksworld, and that ReAct and ToI variants are less successful. The paper also presents ablations on prompting variants and Reflexion, plus an appendix connecting Boomerang to lazy search and posterior sampling.

Significance. The main empirical result, if it survives a controlled comparison, would be practically relevant: Boomerang is simple, cheap in LLM calls, and substantially better than the tested baselines on several domains, and the per-domain tables and released code are useful assets. The weakness is that the central claim about adaptivity is currently confounded with history access, and the numeric inconsistency in the Blocksworld result makes the headline hard to trust as stated. The theoretical appendix does not provide an independent guarantee for Boomerang. With a history-augmented ToI baseline and corrected reporting, the paper could make a solid contribution; in its present form the significance of the adaptivity finding is uncertain.

major comments (4)
  1. [§3.1–3.2, Alg. 1–3, Appendix A.7] The comparison conflates the planning paradigm with access to interaction history. ToI-BFS (Alg. 1) and ToI-DFS (Alg. 3) call the action proposer πθ(s,ϕ,k) and state evaluator Vθ(s,ϕ) on the current state only, whereas Boomerang (Alg. 2) maintains Ht and conditions each new plan on the full history of world-model feedback. Appendix A.7 itself states that ToI makes minimal progress because 'the lack of history in Tree of Interaction (ToI)' discards information. Since a heuristic search framework is not inherently memoryless—one can append the transcript of queried transitions and error messages to the LLM prompts while keeping the same beam/DFS outer loop—the abstract's claim that 'the LLM as a planner can more rapidly adapt its planning strategies based on immediate feedback than LLM as a heuristic' is not established by the current experiments. Please add a history-augmented ToI baseline or restrict the conclusion to the implemented memory-free heuristic.
  2. [§4.2.1–4.2.2 vs. Table 3] There is an unexplained discrepancy in the flagship Blocksworld success rate. The main text reports 0.78 for Boomerang on the 600 PlanBench problems (Fig. 2 and §4.2.1), while Table 3 in Appendix A.10 reports 0.85±0.015 for Boomerang on Blocksworld with average WMQ 10.67. If Table 3 uses a different query budget, seed, or subset, this must be stated; otherwise the headline number is internally inconsistent.
  3. [Appendix A.2, Theorem 1] Theorem 1 is not a bound for Boomerang as implemented. The proof 'follows from Osband et al. (2013)' applies to posterior sampling in reinforcement learning, but §3.2 only conjectures that Boomerang's plan generation approximates posterior sampling from Pθ(M⋆|ϕ,Ht). Without stating that approximation as an explicit assumption (or providing evidence for it), the theorem overclaims a formal guarantee. Please either state the posterior-sampling property as an unproved assumption and label the result conditional, or remove the theorem and treat the connection as intuition.
  4. [§3.2 and §4.1] The query-cost metric assumes free reuse of previously verified transitions. Boomerang re-uses previous queries during verification, so reported WMQ counts unique transitions rather than actual world-model invocations. If the motivating setting is an expensive, non-cacheable world model, the advantage may shrink or reverse. The statement in §4.1 that query efficiency 'directly correlates with time spent in such environments' is not demonstrated. Please report an experiment with an expensive or non-cacheable oracle, or clearly qualify the claim to the cached-query setting.
minor comments (3)
  1. [§4.2.2] The sentence 'Classical uses best-first-search which implies its heuristic does not underestimate the true cost, guaranteeing it finds an optimal path' is incorrect: best-first search does not guarantee optimality, and A* with an admissible heuristic is the usual optimal variant. Given Table 3's configuration, the optimality claim should be corrected.
  2. [§4.2.2] The cross-reference 'see Appendix 3' does not correspond to any appendix section; it should refer to Appendix A.10.
  3. [§4.2.2 and Table 1] The abstract's phrase 'significantly fewer interactions' is not accompanied by statistical tests or confidence intervals for the main 0.78 vs. 0.52/0.63 comparisons; adding error bars or significance tests would strengthen the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the empirical comparisons and the adapted theory are not derived from fitted parameters or from the authors' own prior results.

full rationale

The paper's central empirical claim, that Boomerang is more query-efficient than ToI or classical planners, is measured against external world-model oracles (PlanBench, PDDLGym, and Robotouille) under a fixed query budget, and no parameter is fitted to the reported successes. Boomerang's plans are verified by the true world model, so the success rate is a genuine outcome rather than a quantity forced by construction. The theoretical section explicitly labels the connection to posterior sampling a conjecture and borrows the regret bound from Osband et al. (2013), an external source, so the theory does not independently establish the empirical gains. Self-citations to lazy-search and posterior-sampling literature are motivational and are supported by external, parameter-free proofs, which the paper cites rather than deriving from itself. One experimental-design concern is that ToI's state evaluator receives no interaction history while Boomerang's prompt is updated with the trajectory and error message; this is a potential confound between planning paradigm and memory access, but the paper does not reduce its conclusion to that design by definition. It is a validity threat, not a circular derivation. For these reasons, no circular step is identified and the paper is scored 0.

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

No new physical entities or fitted constants are introduced. The empirical comparison rests on hand-chosen hyperparameters (k, b, temperature, vmin) and on the domain assumptions above; the regret bound rests on a conjectured equivalence between LLM sampling and posterior sampling.

free parameters (4)
  • k (actions proposed per state) for ToI = 2
    Chosen by hand for ToI-BFS and ToI-DFS; increasing k raises world model queries and was not swept.
  • b (beam size) for ToI-BFS = 2
    Chosen by hand; authors report 3, 2, and 2 successes for b=2, 3, 5 on 10 sampled problems, so performance is not highly sensitive but the value still enters the results.
  • temperature = 0.7
    Used for all LLM approaches; affects the diversity of generated plans and replans.
  • vmin (value threshold) for ToI-DFS = not specified
    Algorithm 3 requires this threshold to filter states, but the paper never gives its value, leaving an unspecified hand-tuned parameter in the method.
assumptions (4)
  • domain assumption World model M is deterministic and each query (s,a) returns next state s' and an optional error e.
    Assumed in Section 2; needed so that verified transitions can be cached and re-used across Boomerang iterations.
  • domain assumption Unique world-model queries with caching are the correct cost metric for query-efficient planning.
    Section 3.2 re-uses prior queries; the query-efficiency comparison and the claim that query count correlates with runtime depend on this.
  • ad hoc to paper LLM plan generation approximates posterior sampling over world models for the regret bound.
    Section A.2 conjectures this before Theorem 1; the stated bound for Boomerang requires this assumption, which is not established.
  • domain assumption Classical planner expansions can be counted as world model queries.
    The evaluation counts each FastDownward successor generation as a query so LLM and classical planners can be compared on the same budget.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Query-Efficient Planning with Language Models." pith.science (2026). https://pith.science/paper/JROXP5PK

@misc{pith2026241206162,
  author       = {Pith},
  title        = {Pith review of: Query-Efficient Planning with Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JROXP5PK}},
  note         = {Machine review of arXiv:2412.06162}
}
read the original abstract

Planning in complex environments requires an agent to efficiently query a world model to find a feasible sequence of actions from start to goal. Recent work has shown that Large Language Models (LLMs), with their rich prior knowledge and reasoning capabilities, can potentially help with planning by searching over promising states and adapting to feedback from the world. In this paper, we propose and study two fundamentally competing frameworks that leverage LLMs for query-efficient planning. The first uses LLMs as a heuristic within a search-based planner to select promising nodes to expand and propose promising actions. The second uses LLMs as a generative planner to propose an entire sequence of actions from start to goal, query a world model, and adapt based on feedback. We show that while both approaches improve upon comparable baselines, using an LLM as a generative planner results in significantly fewer interactions. Our key finding is that the LLM as a planner can more rapidly adapt its planning strategies based on immediate feedback than LLM as a heuristic. We present evaluations and ablations on Robotouille and PDDL planning benchmarks and discuss connections to existing theory on query-efficient planning algorithms. Code is available at https://github.com/portal-cornell/llms-for-planning

Figures

Figures reproduced from arXiv: 2412.06162 by the authors.

Figure 1
Figure 1. Overview of LLM planning methods that find a feasible path with minimal queries to a [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Success of approaches that efficiently reached [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Histogram of interactive approaches’ world model queries on Blocksworld problems. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Optimality rate of interactive approaches [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Comparison of ToI-DFS, Boomerang, and Classical planners on Logistics, Grippers, and Robotouille: We chart the success rate given various world model query budgets and observed that Boomerang is most query-efficient at reaching the goal. The applicable classical planne…
Figure 6
Figure 6. Figure 6: Timelapse of a fail￾ure where ReAct repeatedly enters a cul-de-sac when at￾tempting to backtrack (goal in orange) [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 9
Figure 9. Figure 9: Stacked bar plot of successes and fail￾ures modes across I/O variants. ’Invalid Actions’ refers to an outputted action sequence that contains an invalid action. ’Search Failure’ refers to a valid outputted action sequence that does not reach the goal. We break down the…
Figure 10
Figure 10. Figure 10: Timelapse of a failure where ToI-BFS expands various nodes while mak￾ing minimal progress towards the goal [PITH_FULL_IMAGE:figures/full_fig_p026_10.png]
Figure 12
Figure 12. Figure 12: Amount of examples for different numbers of objects in Blocksworld, Logistics, and [PITH_FULL_IMAGE:figures/full_fig_p027_12.png]
Figure 13
Figure 13. Figure 13: Comparison of Reflexion, Boomerang, and ReAct on 100 randomly sampled examples of Blocksworld. We vary the world model query budget and observe Reflexion is able to outperform ReAct but not match Boomerang . Reflexion (Shinn et al., 2023a) produces one action at a tim…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

59 extracted references · 35 canonical work pages

  1. [1]

    Do as i can, not as i say: Grounding language in robotic affordances, 2022

    Michael Ahn, Anthony Brohan, Noah Brown, Yevgen Chebotar, Omar Cortes, Byron David, Chelsea Finn, Chuyuan Fu, Keerthana Gopalakrishnan, Karol Hausman, Alex Herzog, Daniel Ho, Jasmine Hsu, Julian Ibarz, Brian Ichter, Alex Irpan, Eric Jang, Rosario Jauregui Ruano, Kyle Jeffrey, Sally Jesmonth, Nikhil J Joshi, Ryan Julian, Dmitry Kalashnikov, Yuheng Kuang, K...

  2. [2]

    Graph of thoughts: Solving elaborate problems with large language models

    Maciej Besta, Nils Blach, et al. Graph of thoughts: Solving elaborate problems with large language models. arXiv preprint arXiv:2308.09687, 2023

  3. [3]

    Planning as heuristic search

    B Bonet. Planning as heuristic search. Artificial Intelligence, 2001

  4. [4]

    Near-optimal edge evaluation in explicit generalized binomial graphs

    Sanjiban Choudhury, Shervin Javdani, Siddhartha Srinivasa, and Sebastian Scherer. Near-optimal edge evaluation in explicit generalized binomial graphs. In NIPS, 2017

  5. [5]

    Data-driven planning via imitation learning

    Sanjiban Choudhury, Mohak Bhardwaj, Sankalp Arora, Ashish Kapoor, Gireeja Ranade, Sebastian Scherer, and Debadeepta Dey. Data-driven planning via imitation learning. The International Journal of Robotics Research, 37 0 (13-14): 0 1632--1672, 2018

  6. [6]

    Dellin and Siddhartha S

    Christopher M. Dellin and Siddhartha S. Srinivasa. A unifying formalism for shortest path problems with expensive edge evaluations via lazy best-first search over paths with edge selectors. In International Conference on Automated Planning and Scheduling, 2016 a

  7. [7]

    A unifying formalism for shortest path problems with expensive edge evaluations via lazy best-first search over paths with edge selectors

    Christopher M Dellin and Siddhartha S Srinivasa. A unifying formalism for shortest path problems with expensive edge evaluations via lazy best-first search over paths with edge selectors. In ICAPS, 2016 b

  8. [8]

    Guided manipulation planning at the darpa robotics challenge trials

    Christopher M Dellin, Kyle Strabala, G Clark Haynes, David Stager, and Siddhartha S Srinivasa. Guided manipulation planning at the darpa robotics challenge trials. In Experimental Robotics, 2016

Show all 59 references
  1. [9]

    Task and motion planning with large language models for object rearrangement

    Yan Ding, Xiaohan Zhang, Chris Paxton, and Shiqi Zhang. Task and motion planning with large language models for object rearrangement. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 2086--2092. IEEE, 2023

  2. [10]

    Integrated task and motion planning

    Caelan Reed Garrett, Rohan Chitnis, Rachel Holladay, Beomjoon Kim, Tom Silver, Leslie Pack Kaelbling, and Tom \'a s Lozano-P \'e rez. Integrated task and motion planning. Annual review of control, robotics, and autonomous systems, 4: 0 265--293, 2021

  3. [11]

    An empirical analysis of some heuristic features for planning through local search and action graphs

    Alfonso Gerevini, Alessandro Saetti, and Ivan Serina. An empirical analysis of some heuristic features for planning through local search and action graphs. Fundam. Inform., 107 0 (2-3): 0 167--197, 2011. doi:10.3233/FI-2011-399. URL https://doi.org/10.3233/FI-2011-399

  4. [12]

    Critic: Large language models can self-correct with tool-interactive critiquing, 2024

    Zhibin Gou, Zhihong Shao, Yeyun Gong, Yelong Shen, Yujiu Yang, Nan Duan, and Weizhu Chen. Critic: Large language models can self-correct with tool-interactive critiquing, 2024

  5. [13]

    Leveraging pre-trained large language models to construct and utilize world models for model-based task planning

    Lin Guan, Karthik Valmeekam, et al. Leveraging pre-trained large language models to construct and utilize world models for model-based task planning. arXiv preprint arXiv:2305.14909, 2023

  6. [15]

    Reasoning with language model is planning with world model

    Shibo Hao, Yi Gu, et al. Reasoning with language model is planning with world model. arXiv preprint arXiv:2305.14992, 2023 b

  7. [16]

    Lazy collision checking in asymptotically-optimal motion planning

    Kris Hauser. Lazy collision checking in asymptotically-optimal motion planning. In 2015 IEEE international conference on robotics and automation (ICRA), pp.\ 2951--2957. IEEE, 2015

  8. [17]

    The fast downward planning system

    Malte Helmert. The fast downward planning system. Journal of Artificial Intelligence Research, 26: 0 191--246, 2006

  9. [18]

    Posterior sampling for anytime motion planning on graphs with expensive-to-evaluate edges

    Brian Hou, Sanjiban Choudhury, Gilwoo Lee, Aditya Mandalika, and Siddhartha S Srinivasa. Posterior sampling for anytime motion planning on graphs with expensive-to-evaluate edges. In 2020 IEEE International Conference on Robotics and Automation (ICRA), pp.\ 4266--4272. IEEE, 2020

  10. [19]

    Language models as zero-shot planners: Extracting actionable knowledge for embodied agents

    Wenlong Huang, Pieter Abbeel, Deepak Pathak, and Igor Mordatch. Language models as zero-shot planners: Extracting actionable knowledge for embodied agents. In International Conference on Machine Learning, pp.\ 9118--9147. PMLR, 2022 a

  11. [20]

    Inner monologue: Embodied reasoning through planning with language models, 2022 b

    Wenlong Huang, Fei Xia, Ted Xiao, Harris Chan, Jacky Liang, Pete Florence, Andy Zeng, Jonathan Tompson, Igor Mordatch, Yevgen Chebotar, Pierre Sermanet, Noah Brown, Tomas Jackson, Linda Luu, Sergey Levine, Karol Hausman, and Brian Ichter. Inner monologue: Embodied reasoning th...

  12. [21]

    Learning sampling distributions for robot motion planning

    Brian Ichter, James Harrison, and Marco Pavone. Learning sampling distributions for robot motion planning. arXiv preprint arXiv:1709.05448, 2017

  13. [22]

    Offline reinforcement learning as one big sequence modeling problem

    Michael Janner, Qiyang Li, and Sergey Levine. Offline reinforcement learning as one big sequence modeling problem. Advances in neural information processing systems, 34: 0 1273--1286, 2021

  14. [23]

    Integrated task and motion planning in belief space

    Leslie Pack Kaelbling and Tom \'a s Lozano-P \'e rez. Integrated task and motion planning in belief space. The International Journal of Robotics Research, 32 0 (9-10): 0 1194--1227, 2013

  15. [24]

    Llms can't plan, but can help planning in llm-modulo frameworks, 2024

    Subbarao Kambhampati, Karthik Valmeekam, Lin Guan, Kaya Stechly, Mudit Verma, Siddhant Bhambri, Lucas Saldyt, and Anil Murthy. Llms can't plan, but can help planning in llm-modulo frameworks, 2024

  16. [25]

    Learning to search in task and motion planning with streams

    Mohamed Khodeir, Ben Agro, and Florian Shkurti. Learning to search in task and motion planning with streams. IEEE Robotics and Automation Letters, 8 0 (4): 0 1983--1990, 2023

  17. [26]

    Planning algorithms

    Steven M LaValle. Planning algorithms. Cambridge university press, 2006

  18. [27]

    Lee, Annie Xie, Aldo Pacchiano, Yash Chandak, Chelsea Finn, Ofir Nachum, and Emma Brunskill

    Jonathan N. Lee, Annie Xie, Aldo Pacchiano, Yash Chandak, Chelsea Finn, Ofir Nachum, and Emma Brunskill. Supervised pretraining can learn in-context reinforcement learning, 2023

  19. [28]

    Beyond a*: Better planning with transformers via search dynamics bootstrapping, 2024

    Lucas Lehnert, Sainbayar Sukhbaatar, DiJia Su, Qinqing Zheng, Paul Mcvay, Michael Rabbat, and Yuandong Tian. Beyond a*: Better planning with transformers via search dynamics bootstrapping, 2024

  20. [29]

    Text2motion: From natural language instructions to feasible plans

    Kevin Lin, Christopher Agia, Toki Migimatsu, Marco Pavone, and Jeannette Bohg. Text2motion: From natural language instructions to feasible plans. Autonomous Robots, 47 0 (8): 0 1345--1365, 2023

  21. [30]

    Llm+ p: Empowering large language models with optimal planning proficiency

    Bo Liu, Yuqian Jiang, et al. Llm+ p: Empowering large language models with optimal planning proficiency. arXiv preprint arXiv:2304.11477, 2023

  22. [31]

    A constraint-based method for solving sequential manipulation planning problems

    Tom \'a s Lozano-P \'e rez and Leslie Pack Kaelbling. A constraint-based method for solving sequential manipulation planning problems. In 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.\ 3684--3691. IEEE, 2014

  23. [32]

    Smith, and Yejin Choi

    Ximing Lu, Sean Welleck, Peter West, Liwei Jiang, Jungo Kasai, Daniel Khashabi, Ronan Le Bras, Lianhui Qin, Youngjae Yu, Rowan Zellers, Noah A. Smith, and Yejin Choi. Neurologic a*esque decoding: Constrained text generation with lookahead heuristics, 2021

  24. [33]

    Faithful chain-of-thought reasoning, 2023

    Qing Lyu, Shreya Havaldar, Adam Stein, Li Zhang, Delip Rao, Eric Wong, Marianna Apidianaki, and Chris Callison-Burch. Faithful chain-of-thought reasoning, 2023

  25. [34]

    Self-refine: Iterative refinement with self-feedback

    Aman Madaan, Niket Tandon, , et al. Self-refine: Iterative refinement with self-feedback. arXiv preprint arXiv:2303.17651, 2023

  26. [35]

    Srinivasa

    Aditya Mandalika, Oren Salzman, and Siddhartha S. Srinivasa. Lazy receding horizon a* for efficient path planning in graphs with expensive-to-evaluate edges. In International Conference on Automated Planning and Scheduling, 2018

  27. [36]

    Generalized lazy search for robot motion planning: Interleaving search and edge evaluation via event-based toggles

    Aditya Mandalika, Sanjiban Choudhury, Oren Salzman, and Siddhartha Srinivasa. Generalized lazy search for robot motion planning: Interleaving search and edge evaluation via event-based toggles. In Proceedings of the International Conference on Automated Planning and Scheduling...

  28. [37]

    (more) efficient reinforcement learning via posterior sampling, 2013

    Ian Osband, Daniel Russo, and Benjamin Van Roy. (more) efficient reinforcement learning via posterior sampling, 2013

  29. [38]

    Plansformer: Generating symbolic plans using transformers, 2022

    Vishal Pallagani, Bharath Muppasani, Keerthiram Murugesan, Francesca Rossi, Lior Horesh, Biplav Srivastava, Francesco Fabiano, and Andrea Loreggia. Plansformer: Generating symbolic plans using transformers, 2022

  30. [39]

    Heuristics: intelligent search strategies for computer problem solving

    Judea Pearl. Heuristics: intelligent search strategies for computer problem solving. 1984

  31. [41]

    Reflexion: an autonomous agent with dynamic memory and self-reflection

    Noah Shinn, Beck Labash, and Ashwin Gopinath. Reflexion: an autonomous agent with dynamic memory and self-reflection. arXiv preprint arXiv:2303.11366, 2023 b

  32. [42]

    Pddlgym: Gym environments from pddl problems

    Tom Silver and Rohan Chitnis. Pddlgym: Gym environments from pddl problems. In International Conference on Automated Planning and Scheduling (ICAPS) PRL Workshop, 2020. URL https://github.com/tomsilver/pddlgym

  33. [43]

    Tenenbaum, Leslie Pack Kaelbling, and Michael Katz

    Tom Silver, Soham Dan, Kavitha Srinivas, Joshua B. Tenenbaum, Leslie Pack Kaelbling, and Michael Katz. Generalized planning in pddl domains with pretrained large language models, 2023

  34. [44]

    Combined task and motion planning through an extensible planner-independent interface layer

    Siddharth Srivastava, Eugene Fang, Lorenzo Riano, Rohan Chitnis, Stuart Russell, and Pieter Abbeel. Combined task and motion planning through an extensible planner-independent interface layer. In 2014 IEEE international conference on robotics and automation (ICRA), pp.\ 639--6...

  35. [45]

    Adaplanner: Adaptive planning from feedback with language models, 2023

    Haotian Sun, Yuchen Zhuang, Lingkai Kong, Bo Dai, and Chao Zhang. Adaplanner: Adaptive planning from feedback with language models, 2023

  36. [46]

    Differentiable physics and stable modes for tool-use and manipulation planning

    Marc A Toussaint, Kelsey Rebecca Allen, Kevin A Smith, and Joshua B Tenenbaum. Differentiable physics and stable modes for tool-use and manipulation planning. 2018

  37. [47]

    Solving olympiad geometry without human demonstrations

    Trieu Trinh, Yuhuai Tony Wu, Quoc Le, He He, and Thang Luong. Solving olympiad geometry without human demonstrations. Nature, 625: 0 476--482, 2024. URL https://www.nature.com/articles/s41586-023-06747-5

  38. [48]

    On the planning abilities of large language models (a critical investigation with a proposed benchmark), 2023

    Karthik Valmeekam, Sarath Sreedharan, Matthew Marquez, Alberto Olmo, and Subbarao Kambhampati. On the planning abilities of large language models (a critical investigation with a proposed benchmark), 2023

  39. [49]

    Planbench: An extensible benchmark for evaluating large language models on planning and reasoning about change

    Karthik Valmeekam, Matthew Marquez, Alberto Olmo, Sarath Sreedharan, and Subbarao Kambhampati. Planbench: An extensible benchmark for evaluating large language models on planning and reasoning about change. Advances in Neural Information Processing Systems, 36, 2024

  40. [50]

    Demo2code: From summarizing demonstrations to synthesizing code via extended chain-of-thought, 2023

    Huaxiaoyue Wang, Gonzalo Gonzalez-Pumariega, Yash Sharma, and Sanjiban Choudhury. Demo2code: From summarizing demonstrations to synthesizing code via extended chain-of-thought, 2023

  41. [51]

    Yue Wang, Weishi Wang, Shafiq Joty, and Steven C. H. Hoi. Codet5: Identifier-aware unified pre-trained encoder-decoder models for code understanding and generation, 2021

  42. [52]

    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 information processing systems, 35: 0 24824--24837, 2022

  43. [53]

    Self-evaluation guided beam search for reasoning, 2023

    Yuxi Xie, Kenji Kawaguchi, Yiran Zhao, Xu Zhao, Min-Yen Kan, Junxian He, and Qizhe Xie. Self-evaluation guided beam search for reasoning, 2023

  44. [54]

    React: Synergizing reasoning and acting in language models

    Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. React: Synergizing reasoning and acting in language models. arXiv preprint arXiv:2210.03629, 2022

  45. [55]

    Tree of thoughts: Deliberate problem solving with large language models

    Shunyu Yao, Dian Yu, et al. Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601, 2023

  46. [56]

    Tree of thoughts: Deliberate problem solving with large language models

    Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Tom Griffiths, Yuan Cao, and Karthik Narasimhan. Tree of thoughts: Deliberate problem solving with large language models. Advances in Neural Information Processing Systems, 36, 2024

  47. [57]

    Large language models as commonsense knowledge for large-scale task planning

    Zirui Zhao, Wee Sun Lee, and David Hsu. Large language models as commonsense knowledge for large-scale task planning. arXiv preprint arXiv:2305.14078, 2023

  48. [58]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

  49. [59]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...

  50. [60]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...

  51. [61]

    Temporary table just to display results

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.