Pith. sign in

REVIEW 4 major objections 6 minor 4 cited by

EscapeBench: Towards Advancing Creative Intelligence of Language Model Agents

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Language-model agents, even with working memory and chain-of-thought reasoning, make only about 15 percent of the progress needed to finish room-escape puzzles without hints, and a new framework with Foresight and Reflection modules cuts…

desk verdict A solid new benchmark for long-horizon agent exploration, but the 'creativity' label outruns what the metrics actually isolate. read the letter →

arxiv 2412.13549 v2 pith:52YMVX2E submitted 2024-12-18 cs.CL cs.AIcs.LG

Cheng Qian , Peixuan Han , Qinyu Luo , Bingxiang He , Xiusi Chen , Yuji Zhang , Hongyi Du , Jiarui Yao
show 4 more authors
Xiaocheng Yang Denghui Zhang Yunzhu Li Heng Ji
This is my paper · ORCID
classification cs.CLcs.AIcs.LG
keywords creativeintelligencelanguagemodelagentsroomescapegamesbenchmarktooluseimplicitgoaldiscoveryForesightandReflectionlong-horizonreasoning
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

This paper introduces EscapeBench, a suite of 36 room-escape game settings designed to test whether language-model agents can reason creatively in unfamiliar environments, not just follow explicit goals. The central finding is that current models, even when equipped with working memory and Chain-of-Thought reasoning, make only about 15% progress on their own before needing a hint, while average humans reach about 60% before their first hint. To close part of that gap, the paper proposes EscapeAgent, which adds a Foresight module for hypothesizing tool uses before acting and a Reflection module that keeps an explicit task list of unsolved goals and failed attempts. Across ten models, EscapeAgent reduces hint use and total steps by up to 40%, sustains logically coherent action chains of more than 1,000 steps, and improves robustness across difficulty levels. The paper's contribution is a reproducible way to measure creative intelligence in agents and a framework that raises that measured performance.

What carries the argument

The machinery is a graph-structured game engine of scenes, items, and tools with a five-action space (Move, Click, Apply, Input, Craft) and human-annotated 'key steps' that define progress; an agent is scored on Hints Used, Total Steps, and Early Exit Progress. On top of the BaseAgent (working memory plus Chain-of-Thought), EscapeAgent adds two modules: Foresight, which hypothesizes and evaluates tool applications and crafting combinations before acting, and Reflection, which maintains a task list with new/update/delete operations so that failed attempts are recorded and completed goals are removed. The key steps and the hint-after-50-stalls rule are what convert free-form puzzle-solving into a measurable creativity score.

What would settle it

Run EscapeBench under an ablation that replaces Foresight and Reflection with a simple no-repeats memory that records every tried action and forbids retrying it; if that rule-based agent reaches EscapeAgent-level hint-free progress, then the measured gains come from memory management, not creative reasoning.

Watch

Extended reading notes

Core claim

The paper's core claim is that creative reasoning can be operationalized and measured in a text-based agent benchmark, and that current language models fail at it. The evidence is that the best BaseAgent configurations—GPT-4o and Claude-3.5-Sonnet—achieve only about 24–29% early-exit progress (progress made before the first hint), that most hints are consumed on 'key steps' that require creative reasoning, and that 'Input' and 'Craft' actions, the most open-ended, demand the highest relative hint usage. EscapeAgent then shows that these failures are partly addressable: combining Foresight (proposing and evaluating tool-use hypotheses before executing) and Reflection (maintaining an unsolved-task list) improves early-exit progress, reduces hint dependence and steps, and lets agents carry out coherent action chains exceeding 1,000 steps.

Load-bearing premise

The benchmark's definition of creativity is that making progress (hitting a human-annotated key step or collecting a tool) without hints reflects creative reasoning; if that progress is actually gated by memory width, action parsing, or environment-following rather than creativity, the central conclusion about creative intelligence is unsupported.

Editorial extensions

If this is right

  • The 15-percent hint-free progress figure gives a concrete, reproducible baseline: an agent that improves this number without extra hints is demonstrably better at creative problem-solving in unfamiliar environments.
  • EscapeAgent's gains are consistent across closed and open models, with larger models benefiting most, which implies the core model's reasoning capacity remains a binding constraint on creativity.
  • The ablation shows Foresight and Reflection are complementary, so either module alone is a weaker intervention; a full creative-reasoning framework needs both goal tracking and hypothesis generation.
  • Models below about 7B parameters behave close to randomly on this benchmark, establishing a practical scale floor for future agent-creativity evaluations.
  • Domain-specialized models (e.g., math-tuned) can lose creative flexibility, suggesting that over-specialization is measurable in this benchmark.

Reading between the lines

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

  • If progress without hints is accepted as a proxy for creativity, the benchmark invites a direct construct-validity test: vary feedback granularity and memory width while holding game logic fixed and see how much of the human–model gap is attributable to creativity versus memory or instruction-following.
  • The Foresight–Reflection pattern could transfer to other open-ended agent settings, such as web navigation or embodied exploration, where goals are implicit and tool affordances are not memorized.
  • The authors' own data suggest an additional testable extension: giving agents a 'creativity hint' that points to an unusual tool affordance (e.g., 'the stick's hardness') might selectively improve performance, which would verify that the bottleneck is hypothesis generation rather than action execution.
  • A step-reward reinforcement learning variant, suggested in the appendix, would convert EscapeBench from a measurement instrument into a training signal; if step-level rewards improve hint-free progress more than outcome-only rewards, that would support the benchmark's decomposition of creativity into identifiable progress steps.
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 / 6 minor

Summary. EscapeBench constructs 36 text-based room-escape game settings with three difficulty variants and a five-action interface (Move, Click, Apply, Input, Craft), annotated with key steps and linearized solution chains by eight annotators. The paper evaluates ten LMs inside a BaseAgent that has chain-of-thought reasoning and working memory, reports that models make only about 15% average early-exit progress without hints, and then introduces EscapeAgent with Foresight and Reflection modules that reduce hint usage and total steps. The central claim is that these results measure creative intelligence and that current LMs are markedly deficient in it.

Significance. The resource itself is potentially valuable: the environments target genuine open-ended tool use and implicit goal discovery, and the release of code and data plus the detailed appendix documentation are strengths. The multi-model comparison and ablations give a first map of agent performance in this setting. However, the paper's headline quantity, creative reasoning, is not yet cleanly isolated: the absence of a control condition that removes the creative-discovery component, together with the small author-annotator human baseline, means the strong interpretation of the results is not yet established. If the authors add a recipe-following control and a more credible human baseline, the benchmark could become a useful community resource.

major comments (4)
  1. [§5.1, 'Early Exit Progress' definition; Table 3] The load-bearing interpretation that roughly 15% early-exit progress reflects a deficit in creative reasoning is not supported as stated. Progress is defined as achieving a human-annotated key step or collecting a tool, and the hint mechanism in Appendix D.1 gives the next uncompleted action from a linearized solution chain whenever the agent stalls for 50 actions. These design choices mean the metric is entangled with annotator granularity, the linearization of the solution chain, long-horizon credit assignment, and the fixed memory length of 10, in addition to any creative-discovery difficulty. The paper's own error analysis in Table 4 lists environment-following as a separate failure mode. To support the creativity attribution, the authors should add a control condition in which the agent is told the next key step, or given the next target action, at each point, so that discovering the creative step is not required; if such a recipe-following agent still shows low progress or high hint dependence, the bottleneck is execution, not creativity.
  2. [Appendix E; Table 3] The 'Average Human' row is not a fair reference for the human-model gap claimed in §5.3 and Figure 7. Appendix E states that the human baseline is drawn from the eight annotators who authored the game logic and solution chains; they therefore know the solution to each game before playing, and the group is small, homogeneous (U.S.-based CS students), and self-selected for prior room-escape experience. The comparison between this expert/author baseline and a zero-shot LM baseline is not an average-human comparison. The authors should either run a separate user study with naive, independent players, reporting sample size, variance, and completion rates, or explicitly relabel the baseline as an expert/author upper bound and soften claims such as 'humans rarely make mistakes.'
  3. [Tables 3 and 5; §5.3] All model numbers are reported as point estimates from a single run at T=0 with no confidence intervals or repeated seeds. Closed-source APIs are not deterministic in practice, and many of the differences that support the headline reductions, such as hint reductions ranging from 21% to 51% across models and early-exit progress improvements of 0.08 to 22.28 points, are presented without any noise model. The quantitative claims in the abstract and §5.3, including 'achieve only 15% average progress' and 'up to 40% fewer steps and hints,' need at least three repeated runs per model and per condition, or another explicit account of why single-run T=0 evaluation is stable for this environment. This is especially important because the human comparison in Table 3 itself has no variance.
  4. [§4-§5; Appendix C] The paper does not state whether EscapeAgent's prompts, module triggers, and thresholds were tuned or selected using the same 36 game settings and key-step annotations that are later used for evaluation. The Foresight prompts in Appendix C contain substantial task-specific guidance, such as evaluating each tool against the task list and deciding among click, apply, and input, and the hint mechanism uses the same annotation chain. If the framework was iteratively developed on these games, the reported improvement over BaseAgent is an optimistic in-sample estimate. Reporting a development/validation split, or at minimum describing any tuning process and how much the design was adjusted on the test set, is necessary to assess the generality of EscapeAgent.
minor comments (6)
  1. [Abstract; §5.3] The claim of logical coherence over 1,000-step action chains appears in the abstract, but no explicit coherence metric is defined; clarify whether this refers to completion curves, action success rates, or task-list consistency.
  2. [Throughout] Typographical errors should be corrected, including 'EscapAgent' in the Table 5 caption, 'LLama-3.1-70B' in §3.4, 'out main table' in §6.1, 'whcih' in Appendix C, 'acton' in Appendix C.3, and 'shuold' in Appendix C.3.
  3. [Tables 3 and 5] The notation with an arrow plus two numbers, such as '5.03↓5.27,' is ambiguous; state clearly which value is BaseAgent, which is EscapeAgent, and what the arrow denotes, and use consistent formatting with Table 3.
  4. [Table 3; Figure 5] The normalized columns 'Tool Hints Used (percentage)' and 'Key Step Hints Used (percentage)' should specify the normalization denominator in the table caption or text; as printed, the raw count followed by a percentage is easy to misread.
  5. [Appendix D.1] The help instruction says the next target location and next target action will be provided, but it is unclear whether the hint is repeated verbatim at every stalled step or only once; specify the exact help schedule.
  6. [§8] The claim of being 'the first benchmark for LM creativity' is too strong given that the paper itself discusses earlier AUT and TTCT-style evaluations; suggest softening to 'the first agentic room-escape benchmark for LM creativity.'

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the benchmark and agent results are empirical measurements with external yardsticks (human baseline, oracle chain); no load-bearing self-citation or fitted-parameter-as-prediction.

full rationale

EscapeBench is an empirical benchmark and agent evaluation paper, not a derivation. The central numbers (15% average progress, hint counts, step counts, EscapeAgent reductions) are direct measurements from game environments, with an external human baseline and an oracle action chain. The only self-referential flavor is that EscapeAgent is designed and evaluated against the same key-step/tool-collection progress signal used throughout the benchmark, but the paper does not fit parameters to a subset and then predict a closely related quantity; the agent is not trained on the gold annotations, and the prompts it uses do not encode the linearized solution chain. The paper's related-work citations to the authors' own prior publications (e.g., Creator, ToolLink) are contextual and not load-bearing. The operational definition of progress (achieving a human-annotated key step or collecting a tool) is a benchmark construct; concerns about whether that construct measures creativity rather than long-horizon execution or environment-following are construct-validity/correctness issues, not circularity under the criteria here. The Limitations section and Appendix F acknowledge annotation scalability and human-expertise dependence, but these do not make any result reduce by construction to its own input. No step reduces to a fit, a self-citation chain, or a definitional equivalence.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No free parameters beyond human-annotated benchmark definitions; the central claims rest on domain assumptions about what room-escape puzzles measure and on the construct validity of the progress and hint metrics.

assumptions (3)
  • domain assumption Room escape games require creative reasoning, including unconventional tool use and implicit goal discovery.
    Section 3 argues this is the core research question; if false, the benchmark measures something other than creativity.
  • ad hoc to paper Hint usage and total steps are valid quantitative proxies for creative intelligence in this setting.
    Metrics are introduced in Section 5.1 without external validation against established creativity tests such as AUT or TTCT.
  • domain assumption The human baseline is a fair approximation of average human performance.
    Appendix E shows participants were U.S.-based CS students with prior room-escape experience; only 20% self-identified as skilled, yet the paper generalizes to 'average human'.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EscapeBench: Towards Advancing Creative Intelligence of Language Model Agents." pith.science (2026). https://pith.science/paper/52YMVX2E

@misc{pith2026241213549,
  author       = {Pith},
  title        = {Pith review of: EscapeBench: Towards Advancing Creative Intelligence of Language Model Agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/52YMVX2E}},
  note         = {Machine review of arXiv:2412.13549}
}
read the original abstract

Language model agents excel in long-session planning and reasoning, but existing benchmarks primarily focus on goal-oriented tasks with explicit objectives, neglecting creative adaptation in unfamiliar environments. To address this, we introduce EscapeBench, a benchmark suite of room escape game environments designed to challenge agents with creative reasoning, unconventional tool use, and iterative problem-solving to uncover implicit goals. Our results show that current LM models, despite employing working memory and Chain-of-Thought reasoning, achieve only 15% average progress without hints, highlighting their limitations in creativity. To bridge this gap, we propose EscapeAgent, a framework designed to enhance creative reasoning through Foresight (innovative tool use) and Reflection (identifying unsolved tasks). Experiments show that EscapeAgent can execute action chains over 1,000 steps while maintaining logical coherence. It navigates and completes games with up to 40% fewer steps and hints, performs robustly across difficulty levels, and achieves higher action success rates with more efficient and innovative puzzle-solving strategies.

Figures

Figures reproduced from arXiv: 2412.13549 by the authors.

Figure 1
Figure 1. An agent with creative thinking should adapt [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An illustration of Scenes, Tools, and Items in the game and their relations with agent action space. Tools can be collected for “Apply” and “Craft”, while items require “Input”, “Click” or “Apply” of tools to trigger effects [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Statistics of total Scenes, Tools, and Items across all game settings. “Key Steps” refer to the essen￾tial bottleneck actions required to complete the game. • Craft (Tool, Tool): Use two tools in the bag to craft a new one [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: Illustration of the EscapeAgent design. Building on the BaseAgent (A [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Distribution of Key Steps Hints Used, cate [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: The accumulated number of completed games [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Case study on Human, BaseAgent, and EscapeAgent’s progress map corresponding to six game settings. [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Distribution of step intervals for progress made through tool collection and key step achieve￾ment. EscapeAgent uses fewer steps to achieve the next progress and relies less on hints. game scenarios with normal difficulty. Across both GPT-4o and Llama-3.1-70B in [PITH…
Figure 9
Figure 9. Figure 9: More model’s analysis on progress-making interval, extension of Figure [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]
Figure 10
Figure 10. Figure 10: More model’s analysis on item trial times, extension of Figure [PITH_FULL_IMAGE:figures/full_fig_p020_10.png]
Figure 11
Figure 11. Figure 11: An illustration of progress-making trend through all 36 game settings. [PITH_FULL_IMAGE:figures/full_fig_p020_11.png]
Figure 12
Figure 12. Figure 12: Ablation of difficulty through Hints Used. [PITH_FULL_IMAGE:figures/full_fig_p021_12.png]
Figure 13
Figure 13. Figure 13: Ablation of difficulty through Total Steps. [PITH_FULL_IMAGE:figures/full_fig_p021_13.png]
Figure 14
Figure 14. Figure 14: A comparison of action trial times distribu [PITH_FULL_IMAGE:figures/full_fig_p021_14.png]
Figure 15
Figure 15. Figure 15: More model’s analysis on progress with respect to action steps, extension of Figure [PITH_FULL_IMAGE:figures/full_fig_p022_15.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

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

  1. Hell or High Water: Evaluating Agentic Recovery from External Failures

    cs.CL 2025-08 conditional novelty 7.0 of 10

    Language-model agents fail badly at finding backup plans when a planned function is disabled, even when a correct alternative is guaranteed to exist.

  2. UserBench: An Interactive Gym Environment for User-Centric Agents

    cs.AI 2025-07 conditional novelty 6.0 of 10

    A new multi-turn agent benchmark shows that current LLMs elicit fewer than 30% of user preferences and reach full intent alignment only about 20% of the time.

  3. TextAtari: 100K Frames Game Playing with Language Agents

    cs.CL 2025-06 conditional novelty 5.0 of 10

    TextAtari is a text-based Atari benchmark for language agents; 7-8B LLMs stay below 10% of human scores in over 90% of tested conditions, and knowledge injection helps more than chain-of-thought.

  4. Acting Less is Reasoning More! Teaching Model to Act Efficiently

    cs.AI 2025-04 reject novelty 5.0 of 10

    A tool-use efficiency reward for RL-trained LLM agents reduces tool calls by up to 68% and improves tool productivity by up to 215% without significant accuracy loss.

Reference graph

Works this paper leans on

19 extracted references · 11 canonical work pages · cited by 4 Pith papers

  1. [1]

    In this case , use acton ' craft ( < collected tool > , < applicable tool >) ', e

    Combine this tool with another one in your bag to craft a new tool . In this case , use acton ' craft ( < collected tool > , < applicable tool >) ', e . g . craft ( controller , battery ) indicates use a battery in your bag you already have to the controller you just collected to craft a charged controller

    work page

  2. [2]

    In this case , use action ' apply ( < collected tool > , Target Item in a task ) ', e

    Apply this tool to a target item in a task to try solve this task . In this case , use action ' apply ( < collected tool > , Target Item in a task ) ', e . g . apply ( key , locked cabinet ) indicates apply the key you just collected to a locked cabinet to open it . Here are some general hints that you may follow :

    work page

  3. [3]

    Advances in Neural Information Processing Systems, 35:18343– 18362

    Minedojo: Building open-ended embodied agents with internet-scale knowledge. Advances in Neural Information Processing Systems, 35:18343– 18362. Giorgio Franceschelli and Mirco Musolesi. 2023. On the creativity of large language models. arXiv preprint arXiv:2304.00008. Hiroki Furuta, Kuang-Huei Lee, Ofir Nachum, Yutaka Matsuo, Aleksandra Faust, Shixiang S...

    arXiv 2023

  4. [4]

    Journal of Creativity, 33(3):100065

    The originality of machines: Ai takes the torrance test. Journal of Creativity, 33(3):100065. Shibo Hao, Yi Gu, Haodi Ma, Joshua Hong, Zhen Wang, Daisy Wang, and Zhiting Hu. 2023. Rea- soning with language model is planning with world model. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 8154–8173. Chenxu ...

    arXiv 2023

  5. [6]

    In The Twelfth International Conference on Learning Representations

    Smartplay: A benchmark for llms as intelligent agents. In The Twelfth International Conference on Learning Representations. Tianbao Xie, Danyang Zhang, Jixuan Chen, Xiaochuan Li, Siheng Zhao, Ruisheng Cao, Toh Jing Hua, Zhou- jun Cheng, Dongchan Shin, Fangyu Lei, et al. 2024. Osworld: Benchmarking multimodal agents for open- ended tasks in real computer e...

    arXiv 2024

  6. [7]

    Advances in Neural Information Processing Systems, 36

    Tree of thoughts: Deliberate problem solving with large language models. Advances in Neural Information Processing Systems, 36. Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik R Narasimhan, and Yuan Cao. 2023. React: Synergizing reasoning and acting in language models. In The Eleventh International Conference on Learning Representations....

    arXiv 2023

  7. [8]

    hallway”. It leads to nearby scenes including “blocked path close-up

    Assessing and understanding creativity in large language models. Preprint, arXiv:2401.12491. Duo Zheng, Shijia Huang, Lin Zhao, Yiwu Zhong, and Liwei Wang. 2024. Towards learning a generalist model for embodied navigation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13624–13634. Wanjun Zhong, Lianghong Guo, ...

    arXiv 2024

  8. [11]

    Please especially pay attention to the description of the task and the tool , try to find the connection between them to justify your action

    work page

Show all 19 references
  1. [12]

    ' part in response , you shuold explicitly think about whether there 's item in bag for crafting , or task in the list for applying this tool

    In your '- Thought : ... ' part in response , you shuold explicitly think about whether there 's item in bag for crafting , or task in the list for applying this tool . You should read and infer carefully from the tool descriptions and the task description , and evaluate one by one

  2. [13]

    ' part in response , you should give zero to multiple action calls

    In your '- Actions : ... ' part in response , you should give zero to multiple action calls . For each action , you should follow the format ' craft ( < collected tool > , < applicable tool >) ' or ' apply ( < collected tool > , Target Item in a task ) '. If it 's a craft acti...

  3. [14]

    In this case , use action ' click ( Target Item in current task ) ', e

    Click the target item to simply interact with it to solve the task . In this case , use action ' click ( Target Item in current task ) ', e . g . click ( microwave ) indicates click the microwave to examine it and try solve the task

  4. [15]

    In this case , use action ' apply ( < applicable tool > , Target Item in current task ) ', e

    Use the tool in your bag to apply to the target item in the task . In this case , use action ' apply ( < applicable tool > , Target Item in current task ) ', e . g . apply ( key , locked cabinet ) indicates apply the key in your bag to a locked cabinet to open it

  5. [16]

    In this case , use action ' input ( < any string > , Target Item in current task ) ', e

    Input a string to the target item in the task . In this case , use action ' input ( < any string > , Target Item in current task ) ', e . g . input (2413 , combination lock ) indicates input a string password to the combination lock to solve the task . Here are some general hi...

  6. [17]

    Please always first try simple click to interact if haven ' t done so

    Please especially pay attention to the description of the task about what might be needed . Please always first try simple click to interact if haven ' t done so . Examine the tool description and your memory pad , try to find the connection between them and what this task nee...

  7. [18]

    ' part in response , you should explicitly think about whether there 's item to click , tool in bag for applying , or hint from memory pad and tools for string input

    In your '- Thought : ... ' part in response , you should explicitly think about whether there 's item to click , tool in bag for applying , or hint from memory pad and tools for string input . You should read and infer carefully from the task description , evaluate one by one

  8. [19]

    Free Explore

    In your '- Actions : ... ' part in response , you should give zero to multiple action calls . For each action , you should follow the format ' click ( Target Item in current task ) ', ' apply ( < applicable tool > , Target Item in current task ) ', or ' input ( < any string > ...

    2024

  9. [2022]

    In Proceedings of the 31st ACM International Con- ference on Information & Knowledge Management, pages 3023–3032

    Reliable: Offline reinforcement learning for tactical strategies in professional basketball games. In Proceedings of the 31st ACM International Con- ference on Information & Knowledge Management, pages 3023–3032. Xiusi Chen, Wei-Yao Wang, Ziniu Hu, David Reynoso, Kun Jin, Ming...

    2024 arXiv

  10. [2023]

    arXiv preprint arXiv:2304.05332

    Emergent autonomous scientific research ca- pabilities of large language models. arXiv preprint arXiv:2304.05332. Andres M Bran, Sam Cox, Oliver Schilter, Carlo Baldas- sari, Andrew D White, and Philippe Schwaller. 2023. Chemcrow: Augmenting large-language models with chemistr...

    2023 arXiv

  11. [2024]

    arXiv preprint arXiv:2403.05530

    Gemini 1.5: Unlocking multimodal under- standing across millions of tokens of context. arXiv preprint arXiv:2403.05530. Qwen Team. 2024. Qwen2.5: A party of foundation models. Muhtar Ça˘gkan Uluda˘glı and Kaya O˘guz. 2023. Non- player character decision-making in computer game...

    2024 arXiv

Pith tools

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