Pith. sign in

REVIEW 4 major objections 6 minor 97 references

From Words to Workflows: Automating Business Processes

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 Text2Workflow generates business workflows from natural language, hitting 71.3% accuracy and beating single-prompt GPT-4o mini by 27.5% on complex requests.

desk verdict A useful dataset and plausible pipeline, but the headline accuracy advantage is a human-in-the-loop artifact: without feedback the method loses to the baseline. read the letter →

arxiv 2412.03446 v1 pith:S6GPHPPI submitted 2024-12-04 cs.AI

Laura Minkova , Jessica López Espejel , Taki Eddine Toufik Djaidja , Walid Dahhane , El Hassane Ettifouri This is my paper · ORCID
classification cs.AI
keywords NaturalLanguageProcessingGenerativeArtificialIntelligenceLargeModelIntelligentAutomationWorkflowText2WorkflowProcess2JSONJSONgeneration
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 argues that a multi-stage prompt pipeline with user checkpoints can convert natural-language business requests into structured JSON workflows far more reliably than a single large prompt, especially when requests are complex. To test this, the authors built Text2Workflow, a seven-layer prompt system that first screens the request, builds a workflow skeleton, lets a human approve or edit it, then fills in step-level parameters. They also introduce Process2JSON, a 60-request dataset split into easy, medium, and hard levels with expected JSON outputs. On hard requests, Text2Workflow scores 57.5% accuracy against 30% for a single-prompt gpt-4o-mini baseline and 8.8% for gpt-3.5, while overall it reaches 71.3% accuracy. If these numbers hold, the main value is that non-experts could generate executable process blueprints by describing what they want, without writing code.

What carries the argument

The machine that carries the argument is a predefined JSON workflow schema combined with a seven-layer prompt pipeline. The schema separates a general process header (id, name, description, parameters, steps, defaultStartStepId, context) from step-specific structures, with step types such as Decision, Loop, Calculation, DataExtraction, API, Exception, and Unknown. The pipeline uses a Master Prompt to build the workflow skeleton, Expert Prompts to fill parameters step by step, a Parameter Expert for API-like steps, a Questions Prompt to surface missing fields, and a Workflow Modification Prompt to apply user edits. The human feedback loop, inspired by FlowMind, summarizes the skeleton in plain language for user approval or revision before details are filled in. That loop is the main mechanism by which the method improves on hard requests.

What would settle it

Score the same Text2Workflow outputs with multiple independent evaluators and with gold-standard workflows produced by a separate, non-LLM process, then compare inter-rater agreement and absolute scores on the hard subset; if agreement is low or the scores drop materially, the claimed 27.5-point advantage over the gpt-4o-mini baseline would not survive an objective re-measurement.

Watch

Extended reading notes

Core claim

The central claim is that decomposing the workflow-generation task into a sequence of specialized prompts, with a human feedback loop at the skeleton stage, yields substantially more accurate JSON workflows than asking one LLM to do the whole job in a single long prompt. The paper reports that Text2Workflow achieves the highest overall accuracy (71.3%) across all experiments, and that on the hardest requests it surpasses baseline-gpt-3.5-0125 by 48.7 percentage points and baseline-gpt-4o-mini by 27.5 points. The ablation study attributes part of the gain to the human feedback loop, which alone improves accuracy by over 10%, while the logic-screening step adds only a modest 2% improvement.

Load-bearing premise

The reported accuracy numbers assume the expected JSON workflows in Process2JSON are correct and that the semantic scoring rubric, applied by one author-evaluator, is a trustworthy measure of workflow quality; if those gold standards were produced by the same kind of model or are biased toward the authors' own schema, the accuracy comparisons would measure consistency with that format rather than true executability.

Editorial extensions

If this is right

  • If the results replicate, decomposing LLM tasks into expert sub-prompts becomes a viable alternative to fine-tuning for structured-output generation tasks.
  • The public Process2JSON dataset gives other researchers a benchmark to compare natural-language-to-JSON workflow generation on the same easy, medium, and hard split.
  • A business user could describe a process in a few sentences and receive a JSON blueprint ready for visualization and, ultimately, execution by an automation engine with minimal manual intervention.
  • Because the gains concentrate on hard requests, organizations with convoluted, multi-branch processes stand to benefit most, while simple requests are better served by a single prompt for cost and speed.

Reading between the lines

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

  • The paper's reported accuracy is a measure of semantic similarity to gold-standard JSONs, not of whether a generated workflow runs correctly; extending the pipeline to actually execute the generated workflows would be the natural next test, and the authors' own appendix shows a non-executable Loop example.
  • The single-evaluator scoring rubric likely inflates agreement; an independent multi-evaluator study on the same dataset could shift the reported 71.3% figure, especially at the hard level where scoring is most subjective.
  • The token and time costs (roughly twice the input tokens, up to 163 seconds per run) imply that the technique is practical for offline or semi-automated process design but may need optimization before it fits interactive real-time use.
  • The approach's reliance on OpenAI's cloud API brings security and cost constraints; adapting the same prompt decomposition to an open-weight model would show whether the gain comes from the pipeline or from the particular LLM.
Share X Bluesky LinkedIn Reddit HN

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. The paper introduces Text2Workflow, a multi-stage LLM prompting pipeline that converts natural-language business requests into structured JSON workflows. The pipeline comprises seven prompt layers, including a user request clarification step, a master/expert decomposition into steps, a human feedback loop over the workflow skeleton, and a missing-parameter verification stage. The authors also introduce Process2JSON, a 60-request dataset labeled easy/medium/hard, and compare Text2Workflow against two single-prompt baselines (gpt-3.5-turbo-0125 and gpt-4o-mini). They report that Text2Workflow achieves 71.3% overall JSON accuracy versus 64.2% for the gpt-4o-mini baseline, with the largest gap on hard-level requests (57.5% versus 30%). An ablation separates no user aid (NUA), general clarification (GC), human feedback loop (HFL), and the full system.

Significance. If the reported results are sound, the paper makes a useful practical contribution: a modular, prompt-based workflow-generation pipeline with a reusable schema and a new benchmark dataset, plus a transparency note about the single-evaluator limitation. The strengths are the explicit architectural description, the ablation design that isolates the user-input mechanisms, and the candid acknowledgment in Section 6 that the evaluation is biased. However, the quantitative superiority claim is not yet established. The headline hard-level advantage is confounded with human-in-the-loop assistance, the evaluator is a single non-blinded author, no statistical significance or variance is reported, and the gold-standard dataset construction is undocumented. These issues are load-bearing because they directly support the paper's central claim that Text2Workflow outperforms a strong single-prompt baseline on complex requests.

major comments (4)
  1. [Section 5, Table 4] The central claim that Text2Workflow outperforms baseline-gpt-4o-mini on hard-level requests (57.5% vs. 30%) is confounded by the human feedback loop. The ablation shows that without user aid (NUA) the hard-level accuracy is 26.3%, and with only general clarification (GC) it is 27.5%, both below the 30% baseline. Only with the human feedback loop alone (HFL, 32.5%) and especially the full system (57.5%) does the method surpass the baseline. Since the baseline receives no human feedback, the reported advantage cannot be attributed to the multi-layer prompt architecture alone; it may reflect the human corrections injected through the loop. The paper should add a control that gives the baseline the same human feedback (e.g., baseline-gpt-4o-mini with HFL), or otherwise explain the 25-point jump from HFL (32.5%) to the full system (57.5%) in terms of the specific interaction between GC and HFL, rather than attributing it to the pipeline generally.
  2. [Section 5 and Section 6] The accuracy evaluation rests on the semantic rubric in Table 2 applied by a single author-evaluator, with no inter-annotator agreement, no confidence intervals, and no significance tests. The paper acknowledges this in Section 6, but the acknowledgment does not reduce the load-bearing role of the 57.5% vs. 30% comparison: the same author who supplies the human feedback in the full-pipeline runs also assigns the scores, so the evaluation is not blinded. The authors should either provide multiple independent evaluators with agreement statistics, or use an objective metric (e.g., schema validity plus executable-function correctness), and report variance or significance for the difficulty-stratified results.
  3. [Section 4.1] The construction of the Process2JSON gold-standard workflows is not documented. The paper does not state how the expected JSON outputs were produced—whether manually, by LLM, or by a hybrid process—nor whether the gold standard was validated for executability. If the expected JSONs were authored with the same schema and the same style conventions that the prompts enforce, the reported accuracies may measure consistency with the authors' format rather than executable correctness. The paper should describe the annotation process, the schema provenance, the qualifications of the annotator(s), and any checks performed on the gold-standard workflows.
  4. [Section 5, Figure 13] The claim that Text2Workflow 'achieves the highest performance across experiments' is based on an overall average that is dominated by the hard-level gap. At the easy level, baseline-gpt-4o-mini (92.5%) outperforms Text2Workflow (87.5%), and at the medium level they are close (70% versus 68.8%). The paper should present the difficulty-stratified accuracies with per-cell sample sizes (20 per difficulty level) and should temper the overall-performance claim accordingly, since the practical value of the system depends on where the differences are statistically reliable.
minor comments (6)
  1. [Section 3.2] The text lists 'seven distinct layers of prompts' but then summarizes the pipeline into 'five main mechanisms'; the mapping between the seven layers and the five mechanisms is not made explicit and should be clarified.
  2. [Section 5] The abbreviation 'HLF' appears once ('the three ablation models (NUA, GC, HLF)') while the rest of the paper uses 'HFL' for the human feedback loop; the abbreviation should be standardized.
  3. [Section 5] The sentence 'Text2Workflow surpasses baseline-gpt-3.5-0125 by 48.7%, and equally outperforming baseline-gpt-4o-mini (by 27.5%)' is grammatically awkward; also, the percentage differences should be labeled as percentage points to avoid confusion.
  4. [Section 6 vs. Section 5] Section 6 states that Text2Workflow 'sometimes generates incorrect keys in the JSON structure,' while Section 5 states that 'there are no structural errors in the JSON format'; these claims should be harmonized or qualified.
  5. [References] The reference 'OpenIA, 2024' is a typo for OpenAI, and the citation to the structured-outputs guide should be formatted consistently with the other OpenAI citations.
  6. [Appendix D] The text refers to 'figs. D.57 and D.58' for the erroneous output, but the figures are captioned as (1/3), (2/3), and (3/3); the figure numbering appears inconsistent and should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported accuracies are empirical measurements against an author-built dataset, not outputs forced by definition or by a self-citation chain.

full rationale

Text2Workflow is an empirical LLM-prompting pipeline; its central numbers (71.3% overall accuracy, hard-level 57.5% vs 30% for baseline-gpt-4o-mini) are measured by comparing generated JSON to the Process2JSON gold standards using the rubric in Table 2. Nothing in the paper fits a parameter to the evaluation labels and then reports that same quantity as a prediction; the gold-standard JSONs are not fed into the prompt layers, so the outputs are not equal to the inputs by construction. The paper's only self-citation (Espejel et al., 2023) appears in a background sentence about LLM reasoning capabilities and carries no load in the claimed derivation. The authors' reliance on a single evaluator and an author-constructed gold standard (Section 6: "the evaluation method in this study requires further research, as it is biased by reliance on a single evaluator") is a genuine validity and reproducibility limitation, but it is a measurement-quality concern rather than circularity: the scores are not mathematically forced by the rubric. The hard-level superiority of the full Text2Workflow over the baselines is also confounded by the asymmetric human-in-the-loop setup (baselines receive no feedback, while the full system receives clarification plus up to two feedback rounds), but that asymmetry is an experimental-confounding issue, not a derivation that reduces to its own input. Under the strict standard requiring an exhibited equation or self-citation chain that makes the prediction equivalent to its input, no circular step is present.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The ledger is light on free parameters because Text2Workflow is an empirical system, not a fitted model. The one hand-chosen experimental parameter is the cap of two feedback loop iterations. The central claims depend on several domain assumptions about the expressiveness of the JSON schema, the correctness of the self-built gold standard, the reliability of single-evaluator scoring, and the faithfulness of OpenAI models to the prompt screenshots. If any of those assumptions fails, the reported accuracy numbers do not transfer beyond the authors' setup.

free parameters (1)
  • Max human feedback loop iterations = 2
    The authors capped the human feedback loop at two iterations in all experiments (Section 5). Accuracy on complex requests likely changes with this cap, and no sensitivity analysis is reported.
assumptions (4)
  • domain assumption The predefined JSON schema (general process plus step types Decision, Loop, Calculation, DataExtraction, API, Exception, Unknown) can represent a general class of business processes.
    Invoked throughout Section 3.1; the claim of generalized automation depends on the schema being expressive enough for arbitrary workflows, which is not demonstrated.
  • domain assumption The gold-standard workflows in Process2JSON are correct and independent of the system being evaluated.
    Section 4.1 introduces the dataset but does not describe how expected outputs were created; the evaluation gives the system credit for matching these gold standards.
  • domain assumption The semantic scoring rubric in Table 2, applied by one evaluator, produces reliable accuracy numbers.
    Section 4.2 defines the rubric; Section 6 admits reliance on a single evaluator, which makes the assumption load-bearing.
  • domain assumption gpt-4o-mini and gpt-3.5-turbo-0125 follow the prompts faithfully enough for the comparison to reflect the method rather than model quirks.
    All experiments use OpenAI API models with prompt screenshots in Appendix B; no temperature, seed, or repetition settings are reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of From Words to Workflows: Automating Business Processes." pith.science (2026). https://pith.science/paper/S6GPHPPI

@misc{pith2026241203446,
  author       = {Pith},
  title        = {Pith review of: From Words to Workflows: Automating Business Processes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6GPHPPI}},
  note         = {Machine review of arXiv:2412.03446}
}
read the original abstract

As businesses increasingly rely on automation to streamline operations, the limitations of Robotic Process Automation (RPA) have become apparent, particularly its dependence on expert knowledge and inability to handle complex decision-making tasks. Recent advancements in Artificial Intelligence (AI), particularly Generative AI (GenAI) and Large Language Models (LLMs), have paved the way for Intelligent Automation (IA), which integrates cognitive capabilities to overcome the shortcomings of RPA. This paper introduces Text2Workflow, a novel method that automatically generates workflows from natural language user requests. Unlike traditional automation approaches, Text2Workflow offers a generalized solution for automating any business process, translating user inputs into a sequence of executable steps represented in JavaScript Object Notation (JSON) format. Leveraging the decision-making and instruction-following capabilities of LLMs, this method provides a scalable, adaptable framework that enables users to visualize and execute workflows with minimal manual intervention. This research outlines the Text2Workflow methodology and its broader implications for automating complex business processes.

Figures

Figures reproduced from arXiv: 2412.03446 by the authors.

Figure 1
Figure 1. General Process JSON. The id is a process-unique, random universal unique identifier (UUID). The name key stores a short, couple-word name that summarizes the pro￾cess. The description is a longer, one to two sentence long summary of the process. The parameters key stores process unique parameters, though it is often left empty. The steps key stores a list of individual, unitary steps required to accomplish the proc… view at source ↗
Figure 2
Figure 2. The process of breaking down a natural language user request into a workflow [PITH_FULL_IMAGE:figures/full_fig_p024_2.png] view at source ↗
Figure 3
Figure 3. The user feedback loop mechanism, enabling the user to validate and/or modify [PITH_FULL_IMAGE:figures/full_fig_p024_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Complete workflow generation, including verifying and creating natural sounding [PITH_FULL_IMAGE:figures/full_fig_p025_4.png]
Figure 5
Figure 5. Figure 5: Manually completing the workflow by allowing the user to fill in the missing [PITH_FULL_IMAGE:figures/full_fig_p025_5.png]
Figure 6
Figure 6. Figure 6: Workflow modification mechanism. 4. Evaluation methodology This section presents a thorough analysis of the metrics and evaluation criteria applied to our dataset. 4.1. Proposed Dataset We constructed our own dataset, Process2JSON, consisting of 60 concise example user…
Figure 7
Figure 7. Figure 7: Expected Output JSON for Easy Sample. 26 [PITH_FULL_IMAGE:figures/full_fig_p028_7.png]
Figure 8
Figure 8. Figure 8: Average token usage, across all experiments. [PITH_FULL_IMAGE:figures/full_fig_p032_8.png]
Figure 9
Figure 9. Figure 9: Input token usage, across all experiments and by difficulty level. [PITH_FULL_IMAGE:figures/full_fig_p034_9.png]
Figure 10
Figure 10. Figure 10: Completion token generation, across all experiments and by difficulty level. [PITH_FULL_IMAGE:figures/full_fig_p035_10.png]
Figure 11
Figure 11. Figure 11: Execution time (in seconds), across all experiments. [PITH_FULL_IMAGE:figures/full_fig_p036_11.png]
Figure 12
Figure 12. Figure 12: The JSON accuracy results (in %), by difficulty level. [PITH_FULL_IMAGE:figures/full_fig_p038_12.png]
Figure 13
Figure 13. Figure 13: The overall performances of every method (in %). [PITH_FULL_IMAGE:figures/full_fig_p039_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

97 extracted references · 27 canonical work pages

  1. [1]

    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 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

  2. [2]

    , author Bichler, M

    author van der Aalst, W.M. , author Bichler, M. , author Heinzl, A. , year 2018 . title Robotic process automation . journal Business & Information Systems Engineering volume 60 , pages 269--272 . https://doi.org/10.1007/s12599-018-0542-4, :10.1007/s12599-018-0542-4

    doi:10.1007/s12599-018-0542-4 2018

  3. [3]

    , year 2024

    author Agostinelli, S. , year 2024 . title Limitations, future works and impact , in: booktitle Generating Executable Robotic Process Automation Scripts from Unsegmented User Interface Logs . publisher Springer , pp. pages 101--104

    2024

  4. [4]

    , author Brohan, A

    author Ahn, M. , author Brohan, A. , author Brown, N. , author Chebotar, Y. , author Cortes, O. , author David, B. , author Finn, C. , author Fu, C. , author Gopalakrishnan, K. , author Hausman, K. , author et al. , year 2022 . title Do as i can, not as i say: Grounding language in robotic affordances . https://arxiv.org/abs/2204.01691, http://arxiv.org/a...

    arXiv 2022

  5. [5]

    title Introducing the next generation of claude

    author Anthropic , year 2024 . title Introducing the next generation of claude . howpublished https://www.anthropic.com/news/claude-3-family

    2024

  6. [6]

    , author Singh, S

    author Arora, R. , author Singh, S. , author Swaminathan, K. , author Datta, A. , author Banerjee, S. , author Bhowmick, B. , author Jatavallabhula, K. , author Sridharan, M. , author Krishna, M. , year 2024 a. title Anticipate & act: Integrating llms and classical planning for efficient task execution in household environments. in the ieee international ...

    2024

  7. [7]

    , author Singh, S

    author Arora, R. , author Singh, S. , author Swaminathan, K. , author Datta, A. , author Banerjee, S. , author Bhowmick, B. , author Jatavallabhula, K.M. , author Sridharan, M. , author Krishna, M. , year 2024 b. title Anticipate & act: Integrating llms and classical planning for efficient task execution in household environments . journal International C...

    2024

  8. [8]

    , author Gopalan, A

    author Banerjee, D. , author Gopalan, A. , year 2024 . title Towards reliable alignment: Uncertainty-aware rlhf . https://arxiv.org/abs/2410.23726, http://arxiv.org/abs/2410.23726 arXiv:2410.23726

    arXiv 2024

Show all 97 references
  1. [9]

    , author Cahyawijaya, S

    author Bang, Y. , author Cahyawijaya, S. , author Lee, N. , author Dai, W. , author Su, D. , author Wilie, B. , author Lovenia, H. , author Ji, Z. , , author Yu, T. , author Chung, W. , author Do, Q.V. , author Xu, Y. , author Fung, P. , year 2023 . title A multitask, multilin...

    2023 arXiv

  2. [10]

    , author Yang, J

    author Beheshti, A. , author Yang, J. , author Sheng, Q.Z. , author Benatallah, B. , author Casati, F. , author Dustdar, S. , author Nezhad, H.R.M. , author Zhang, X. , author Xue, S. , year 2023 . title Processgpt: Transforming business process management with generative arti...

    2023

  3. [11]

    , author Gebru, T

    author Bender, E.M. , author Gebru, T. , author McMillan-Major, A. , author Shmitchell, S. , year 2021 . title On the dangers of stochastic parrots: Can language models be too big? , in: booktitle Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transpar...

    2021

  4. [12]

    , author El-Sayed, H

    author Bengesi, S. , author El-Sayed, H. , author Sarker, M.K. , author Houkpati, Y. , author Irungu, J. , author Oladunni, T. , year 2024 . title Advancements in generative ai: A comprehensive review of gans, gpt, autoencoders, diffusion model, and transformers . journal IEEE...

    2024

  5. [13]

    , author Mann, B

    author Brown, T.B. , author Mann, B. , author Ryder, N. , author Subbiah, M. , author Kaplan, J. , author Dhariwal, P. , author Neelakantan, A. , author Shyam, P. , author Sastry, G. , author Askell, A. , year 2020 . title Language models are few-shot learners . journal CoRR v...

    2020 arXiv

  6. [14]

    , author Tworek, J

    author Chen, M. , author Tworek, J. , author Jun, H. , author Yuan, Q. , author de Oliveira Pinto, H.P. , author Kaplan, J. , author Edwards, H. , author Burda, Y. , author Joseph, N. , author et al. , year 2021 . title Evaluating large language models trained on code . https:...

    2021 arXiv

  7. [15]

    , year 2023

    author Chen, W. , year 2023 . title Large language models are few(1)-shot table reasoners . http://arxiv.org/abs/2210.06710 arXiv:2210.06710

    2023 arXiv

  8. [16]

    , author Xia, K

    author Cheong, I. , author Xia, K. , author Feng, K.K. , author Chen, Q.Z. , author Zhang, A.X. , year 2024 . title (a) i am not a lawyer, but...: Engaging legal experts towards responsible llm policies for legal advice , in: booktitle The 2024 ACM Conference on Fairness, Acco...

    2024

  9. [17]

    , author van Merrienboer, B

    author Cho, K. , author van Merrienboer, B. , author Bahdanau, D. , author Bengio, Y. , year 2014 . title On the properties of neural machine translation: Encoder-decoder approaches . https://arxiv.org/abs/1409.1259, http://arxiv.org/abs/1409.1259 arXiv:1409.1259

    2014 arXiv

  10. [18]

    , author Narang, S

    author Chowdhery, A. , author Narang, S. , author Devlin, J. , author Bosma, M. , author Mishra, G. , author Roberts, A. , author Barham, P. , author Chung, H.W. , author Sutton, C. , author et al., S.G. , year 2022 . title Palm: Scaling language modeling with pathways . https...

    2022 arXiv

  11. [19]

    , author Hislop, D

    author Coombs, C. , author Hislop, D. , author Taneva, S.K. , author Barnard, S. , year 2020 . title The strategic impacts of intelligent automation for knowledge and service work: An interdisciplinary review . journal The Journal of Strategic Information Systems volume 29 , p...

    2020

  12. [20]

    , author Mamede, H.S

    author Costa, D.A.d.S. , author Mamede, H.S. , author Silva, M.M.d. , year 2022 . title Robotic process automation (rpa) adoption: a systematic literature review . journal Engineering Management in Production and Services volume 14

    2022

  13. [21]

    , author Chang, M

    author Devlin, J. , author Chang, M. , author Lee, K. , author Toutanova, K. , year 2018 . title BERT: pre-training of deep bidirectional transformers for language understanding . journal CoRR volume abs/1810.04805 . http://arxiv.org/abs/1810.04805, http://arxiv.org/abs/1810.0...

    2018 arXiv

  14. [22]

    , author Jauhri, A

    author Dubey, A. , author Jauhri, A. , author Pandey, A. , author Kadian, A. , author Al-Dahle, A. , author Letman, A. , author Mathur, A. , author Schelten, A. , author Yang, A. , author Fan, A. , year 2024 . title The llama 3 herd of models . https://arxiv.org/abs/2407.21783...

    2024 arXiv

  15. [23]

    , author Ettifouri, E.H

    author Espejel, J.L. , author Ettifouri, E.H. , author Alassan, M.S.Y. , author Chouham, E.M. , author Dahhane, W. , year 2023 . title Gpt-3.5, gpt-4, or bard? evaluating llms reasoning ability in zero-shot setting and performance boosting through prompts . journal Natural Lan...

    2023

  16. [24]

    , author Waddoups, N

    author Eulerich, M. , author Waddoups, N. , author Wagener, M. , author Wood, D.A. , year 2024 . title The dark side of robotic process automation (rpa): Understanding risks and challenges with rpa . journal Accounting horizons volume 38 , pages 143--152

    2024

  17. [25]

    , author Mohd Rahim, M.S

    author Farooq, U. , author Mohd Rahim, M.S. , author Abid, A. , year 2023 . title A multi-stack rnn-based neural machine translation model for english to pakistan sign language translation . journal Neural Computing and Applications volume 35 , pages 13225--13238

    2023

  18. [26]

    , author Qiang, J

    author Feng, Y. , author Qiang, J. , author Li, Y. , author Yuan, Y. , author Zhu, Y. , year 2023 . title Sentence simplification via large language models . https://arxiv.org/abs/2302.11957, http://arxiv.org/abs/2302.11957 arXiv:2302.11957

    2023 arXiv

  19. [27]

    , author Pouget-Abadie, J

    author Goodfellow, I.J. , author Pouget-Abadie, J. , author Mirza, M. , author Xu, B. , author Warde-Farley, D. , author Ozair, S. , author Courville, A. , author Bengio, Y. , year 2014 . title Generative adversarial networks . https://arxiv.org/abs/1406.2661, http://arxiv.org...

    2014 arXiv

  20. [28]

    title Ai achieves silver-medal standard solving international mathematical olympiad problems

    author Google , year 2024 a. title Ai achieves silver-medal standard solving international mathematical olympiad problems . howpublished https://deepmind.google/discover/blog/ai-solves-imo-problems-at-silver-medal-level/ . note Online; accessed 11 October 2024

    2024

  21. [29]

    title Alphageometry: An olympiad-level ai system for geometry

    author Google , year 2024 b. title Alphageometry: An olympiad-level ai system for geometry . howpublished https://deepmind.google/discover/blog/alphageometry-an-olympiad-level-ai-system-for-geometry/ . note Online; accessed 11 October 2024

    2024

  22. [30]

    , author Have, H.t

    author Gordijn, B. , author Have, H.t. , year 2023 . title Chatgpt: evolution or revolution? journal Medicine, Health Care and Philosophy volume 26 , pages 1--2

    2023

  23. [31]

    , author Ranisch, R

    author Haltaufderheide, J. , author Ranisch, R. , year 2024 . title The ethics of chatgpt in medicine and healthcare: a systematic review on large language models (llms) . journal NPJ digital medicine volume 7 , pages 183

    2024

  24. [32]

    , author Jain, A

    author Ho, J. , author Jain, A. , author Abbeel, P. , year 2020 . title Denoising diffusion probabilistic models . journal CoRR volume abs/2006.11239 . https://arxiv.org/abs/2006.11239, http://arxiv.org/abs/2006.11239 arXiv:2006.11239

    2020 arXiv

  25. [33]

    , author Schmidhuber, J

    author Hochreiter, S. , author Schmidhuber, J. , year 1997 . title Long short-term memory . journal Neural Computation volume 9 , pages 1735--1780 . https://doi.org/10.1162/neco.1997.9.8.173

    1997 doi

  26. [34]

    , author Borgeaud, S

    author Hoffmann, J. , author Borgeaud, S. , author Mensch, A. , author Buchatskaya, E. , author Cai, T. , author Rutherford, E. , author de Las Casas, D. , author Hendricks, L.A. , author Welbl, J. , author Clark, A. , author et al. , year 2022 . title Training compute-optimal...

    2022 arXiv

  27. [35]

    , author Castro, I

    author Hosseini, P. , author Castro, I. , author Ghinassi, I. , author Purver, M. , year 2024 . title Efficient solutions for an intriguing failure of llms: Long context window does not mean llms can analyze long sequences flawlessly . journal arXiv preprint arXiv:2408.01866

    2024 arXiv

  28. [36]

    , author Niu, Y

    author Hou, Z. , author Niu, Y. , author Du, Z. , author Zhang, X. , author Liu, X. , author Zeng, A. , author Zheng, Q. , author Huang, M. , author Wang, H. , author Tang, J. , author Dong, Y. , year 2024 . title Chatglm-rlhf: Practices of aligning large language models with ...

    2024 arXiv

  29. [37]

    , author Zhao, P

    author Hu, M. , author Zhao, P. , author Xu, C. , author Sun, Q. , author Lou, J. , author Lin, Q. , author Luo, P. , author Rajmohan, S. , author Zhang, D. , year 2024 . title Agentgen: Enhancing planning abilities for large language model based agent via environment and task...

    2024 arXiv

  30. [38]

    , author Petrovi \'c , M

    author Joki \'c , A. , author Petrovi \'c , M. , author Miljkovi \'c , Z. , year 2023 . title The framework for mobile robot task planning based on the optimal manufacturing schedule , in: booktitle International Workshop on Medical and Service Robots , organization Springer ....

    2023

  31. [39]

    , author Kamigaito, H

    author Juseon-Do, J.D. , author Kamigaito, H. , author Okumura, M. , author Kwon, J. , year 2024 . title Instructcmp: Length control in sentence compression through instruction-based large language models , in: booktitle Findings of the Association for Computational Linguistic...

    2024

  32. [40]

    , author He, S

    author Kang, L. , author He, S. , author Wang, M. , author Long, F. , author Su, J. , year 2023 . title Bilingual attention based neural machine translation . journal Applied Intelligence volume 53 , pages 4302--4315

    2023

  33. [41]

    , author Venkatesh, V.L.N

    author Kannan, S.S. , author Venkatesh, V.L.N. , author Min, B.C. , year 2024 . title Smart-llm: Smart multi-agent robot task planning using large language models . https://arxiv.org/abs/2309.10062, http://arxiv.org/abs/2309.10062 arXiv:2309.10062

    2024 arXiv

  34. [42]

    , author Welling, M

    author Kingma, D.P. , author Welling, M. , year 2022 . title Auto-encoding variational bayes . https://arxiv.org/abs/1312.6114, http://arxiv.org/abs/1312.6114 arXiv:1312.6114

    2022 arXiv

  35. [43]

    , author Gu, S.S

    author Kojima, T. , author Gu, S.S. , author Reid, M. , author Matsuo, Y. , author Iwasawa, Y. , year 2023 . title Large language models are zero-shot reasoners . https://arxiv.org/abs/2205.11916, http://arxiv.org/abs/2205.11916 arXiv:2205.11916

    2023 arXiv

  36. [44]

    , author Rouvier, M

    author Labrak, Y. , author Rouvier, M. , author Dufour, R. , year 2024 . title A zero-shot and few-shot study of instruction-finetuned large language models applied to clinical and biomedical tasks . https://arxiv.org/abs/2307.12114, http://arxiv.org/abs/2307.12114 arXiv:2307.12114

    2024 arXiv

  37. [45]

    , author Finnigan, S

    author Leiker, D. , author Finnigan, S. , author Gyllen, A.R. , author Cukurova, M. , year 2023 . title Prototyping the use of large language models (llms) for adult learning content creation at scale . journal arXiv preprint arXiv:2306.01815

    2023 arXiv

  38. [46]

    , author Zhang, G

    author Li, T. , author Zhang, G. , author Do, Q.D. , author Yue, X. , author Chen, W. , year 2024 . title Long-context llms struggle with long in-context learning . journal arXiv preprint arXiv:2404.02060

    2024 arXiv

  39. [47]

    , author Choi, D

    author Li, Y. , author Choi, D. , author Chung, J. , author Kushman, N. , author Schrittwieser, J. , author Leblond, R. , author Eccles, T. , author Keeling, J. , author Gimeno, F. , author Dal Lago, A. , year 2022 . title Competition-level code generation with alphacode . jou...

    2022 doi

  40. [48]

    , author Xia, C.S

    author Liu, J. , author Xia, C.S. , author Wang, Y. , author Zhang, L. , year 2023 a. title Is your code generated by chatgpt really correct? rigorous evaluation of large language models for code generation . https://arxiv.org/abs/2305.01210, http://arxiv.org/abs/2305.01210 ar...

    2023 arXiv

  41. [49]

    , author Singh, A

    author Liu, Y. , author Singh, A. , author Freeman, C.D. , author Co-Reyes, J.D. , author Liu, P.J. , year 2023 b. title Improving large language model fine-tuning for solving math problems . https://arxiv.org/abs/2310.10047, http://arxiv.org/abs/2310.10047 arXiv:2310.10047

    2023 arXiv

  42. [50]

    , author Yang, C.M

    author Lo, W. , author Yang, C.M. , author Zhang, Q. , author Li, M. , year 2024 . title Increased productivity and reduced waste with robotic process automation and generative ai-powered ioe services . journal Journal of Web Engineering volume 23 , pages 53--87 . :10.13052/jw...

    2024

  43. [51]

    , author Topol, E.J

    author Mesk \'o , B. , author Topol, E.J. , year 2023 . title The imperative for regulatory oversight of large language models (or generative ai) in healthcare . journal NPJ digital medicine volume 6 , pages 120

    2023

  44. [52]

    title Microsoft Copilot Studio

    author Microsoft , year 2024 . title Microsoft Copilot Studio . howpublished https://www.microsoft.com/fr-fr/microsoft-copilot/microsoft-copilot-studio

    2024

  45. [53]

    , author Mikolov, T

    author Minaee, S. , author Mikolov, T. , author Nikzad, N. , author Chenaghlu, M. , author Socher, R. , author Amatriain, X. , author Gao, J. , year 2024 . title Large language models: A survey . journal arXiv preprint arXiv:2402.06196

    2024 arXiv

  46. [54]

    , author Tong, R

    author Moore, S. , author Tong, R. , author Singh, A. , author Liu, Z. , author Hu, X. , author Lu, Y. , author Liang, J. , author Cao, C. , author Khosravi, H. , author Denny, P. , et al., year 2023 . title Empowering education with llms-the next-gen interface and content gen...

    2023

  47. [55]

    , author Mamede, H.S

    author Moreira, S. , author Mamede, H.S. , author Santos, A. , year 2023 . title Process automation using rpa – a literature review . journal Procedia Computer Science volume 219 , pages 244--254 . https://www.sciencedirect.com/science/article/pii/S1877050923002958, :https://d...

    2023 doi

  48. [56]

    , author Karamardian, D

    author Nay, J.J. , author Karamardian, D. , author Lawsky, S.B. , author Tao, W. , author Bhat, M. , author Jain, R. , author Lee, A.T. , author Choi, J.H. , author Kasai, J. , year 2024 . title Large language models as tax attorneys: a case study in legal capabilities emergen...

    2024

  49. [57]

    , author Pang, B

    author Nijkamp, E. , author Pang, B. , author Hayashi, H. , author Tu, L. , author Wang, H. , author Zhou, Y. , author Savarese, S. , author Xiong, C. , year 2023 . title Codegen: An open large language model for code with multi-turn program synthesis . https://arxiv.org/abs/2...

    2023 arXiv

  50. [58]

    title Introducing chatgpt

    author OpenAI , year 2022 . title Introducing chatgpt . howpublished https://openai.com/blog/chatgpt

    2022

  51. [59]

    title Models - openai api

    author OpenAI , year 2023 . title Models - openai api . howpublished https://platform.openai.com/docs/models/gpt-3-5-turbo

    2023

  52. [60]

    title Gpt-4o mini: advancing cost-efficient intelligence

    author OpenAI , year 2024 . title Gpt-4o mini: advancing cost-efficient intelligence . howpublished https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/

    2024

  53. [61]

    , author Adler, S

    author OpenAI , author Achiam, J. , author Adler, S. , author Agarwal, S. , author Ahmad, L. , author Akkaya, I. , author Aleman, F.L. , author Almeida, D. , author Altenschmidt, J. , author Altman, S. , author et al. , year 2024 . title Gpt-4 technical report . https://arxiv....

    2024 arXiv

  54. [62]

    title Structured outputs

    author OpenIA , year 2024 . title Structured outputs . howpublished https://platform.openai.com/docs/guides/structured-outputs/introduction

    2024

  55. [63]

    , author Wu, J

    author Ouyang, L. , author Wu, J. , author Jiang, X. , author Almeida, D. , author Wainwright, C.L. , author Mishkin, P. , author Zhang, C. , author Agarwal, S. , author Slama, K. , author Ray, A. , author et al. , year 2022 . title Training language models to follow instructi...

    2022 arXiv

  56. [64]

    , author Holia, M

    author Pandya, K. , author Holia, M. , year 2023 . title Automating customer service using langchain: Building custom open-source gpt chatbot for organizations . journal arXiv preprint arXiv:2310.05421

    2023 arXiv

  57. [65]

    , author Xie, S

    author Peebles, W. , author Xie, S. , year 2023 . title Scalable diffusion models with transformers , in: booktitle Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. pages 4195--4205

    2023

  58. [66]

    , author Adithyan, R

    author Pinnaparaju, N. , author Adithyan, R. , author Phung, D. , author Tow, J. , author Baicoianu, J. , author Datta, A. , author Zhuravinskyi, M. , author Mahan, D. , author Bellagente, M. , author Riquelme, C. , author Cooper, N. , year 2024 . title Stable code technical r...

    2024 arXiv

  59. [67]

    , author Dolantina, H

    author Qordia, G. , author Dolantina, H. , year 2024 . title The benefits of using ipa in relation to rpa for the cryptocurrency sector, in making decisions on their sale and purchase in the stock market . journal Smart Cities and Regional Development (SCRD) Journal volume 8 ,...

    2024

  60. [68]

    , author Narasimhan, K

    author Radford, A. , author Narasimhan, K. , author Salimans, T. , author Sutskever, I. , year 2018 a. title Improving language understanding by generative pre-training . journal arxiv

    2018

  61. [69]

    , author Wu, J

    author Radford, A. , author Wu, J. , author Child, R. , author Luan, D. , author Amodei, D. , author Sutskever, I. , year 2018 b. title Language models are unsupervised multitask learners . journal arxiv https://d4mucfpksywv.cloudfront.net/better-language-models/language-models.pdf

    2018

  62. [70]

    , author Gehring, J

    author Rozière, B. , author Gehring, J. , author Gloeckle, F. , author Sootla, S. , author Gat, I. , author Tan, X.E. , author Adi, Y. , author Liu, J. , author et al. , year 2024 . title Code llama: Open foundation models for code . http://arxiv.org/abs/2308.12950 arXiv:2308.12950

    2024 arXiv

  63. [71]

    , author Chen, Y

    author Ruan, J. , author Chen, Y. , author Zhang, B. , author Xu, Z. , author Bao, T. , author du qing , author shi shiwei , author Mao, H. , author Zeng, X. , author Zhao, R. , year 2023 . title TPTU : Task planning and tool usage of large language model-based AI agents . jou...

    2023

  64. [72]

    , author Hinton, G.E

    author Rumelhart, D.E. , author Hinton, G.E. , author Williams, R.J. , year 1986 . title Learning representations by back-propagating errors . journal Nature volume 323 , pages 533--536 . https://api.semanticscholar.org/CorpusID:205001834

    1986

  65. [73]

    , author Singh, A.K

    author Sahoo, P. , author Singh, A.K. , author Saha, S. , author Jain, V. , author Mondal, S. , author Chadha, A. , year 2024 . title A systematic survey of prompt engineering in large language models: Techniques and applications . https://arxiv.org/abs/2402.07927, http://arxi...

    2024 arXiv

  66. [74]

    , author Li, J

    author Shi, J. , author Li, J. , author Ma, Q. , author Yang, Z. , author Ma, H. , author Li, L. , year 2024 . title Chops: Chat with customer profile systems for customer service with llms . journal arXiv preprint arXiv:2404.01343

    2024 arXiv

  67. [75]

    , author Aunimo, L

    author Siderska, J. , author Aunimo, L. , author Sube, T. , author von Stamm, J. , author Kedziora, D. , author Aini, S.N.B.M. , year 2023 . title Towards intelligent automation (ia): Literature review on the evolution of robotic process automation (rpa), its challenges, and f...

    2023 doi

  68. [76]

    , author Zhao, H

    author Team, C. , author Zhao, H. , author Hui, J. , author Howland, J. , author Nguyen, N. , author Zuo, S. , author Hu, A. , author Choquette-Choo, C.A. , author Shen, J. , author Kelley, J. , author et al. , year 2024 a. title Codegemma: Open code models based on gemma . ht...

    2024 arXiv

  69. [77]

    , author Anil, R

    author Team, G. , author Anil, R. , author Borgeaud, S. , author Alayrac, J.B. , author Yu, J. , author Soricut, R. , author Schalkwyk, J. , author Dai, A.M. , author Hauth, A. , author Millican, K. , author et al., D.S. , year 2024 b. title Gemini: A family of highly capable ...

    2024 arXiv

  70. [78]

    , author Lavril, T

    author Touvron, H. , author Lavril, T. , author Izacard, G. , author Martinet, X. , author Lachaux, M.A. , author Lacroix, T. , author Rozière, B. , author Goyal, N. , author Hambro, E. , author Azhar, F. , author Rodriguez, A. , author Joulin, A. , author Grave, E. , author L...

    2023 arXiv

  71. [79]

    , author Martin, L

    author Touvron, H. , author Martin, L. , author Stone, K. , author Albert, P. , author Almahairi, A. , author Babaei, Y. , author Bashlykov, N. , author Batra, S. , author Bhargava, P. , author Bhosale, S. , author et al. , year 2023 b. title Llama 2: Open foundation and fine-...

    2023 arXiv

  72. [80]

    , author Suresh, T

    author Ugare, S. , author Suresh, T. , author Kang, H. , author Misailovic, S. , author Singh, G. , year 2024 . title Syncode: Llm generation with grammar augmentation . https://arxiv.org/abs/2403.01632, http://arxiv.org/abs/2403.01632 arXiv:2403.01632

    2024 arXiv

  73. [81]

    title Meet UiPath Autopilot™, your partner for AI at work

    author UiPath , year 2024 . title Meet UiPath Autopilot™, your partner for AI at work . howpublished https://www.uipath.com/product/autopilot

    2024

  74. [82]

    , author Mandlekar, A

    author Urain, J. , author Mandlekar, A. , author Du, Y. , author Shafiullah, M. , author Xu, D. , author Fragkiadaki, K. , author Chalvatzaki, G. , author Peters, J. , year 2024 . title Deep generative models in robotics: A survey on learning from multimodal demonstrations . h...

    2024 arXiv

  75. [83]

    , author Marquez, M

    author Valmeekam, K. , author Marquez, M. , author Sreedharan, S. , author Kambhampati, S. , year 2023 a. title On the planning abilities of large language models - a critical investigation . journal Thirty-seventh Conference on Neural Information Processing Systems https://op...

    2023

  76. [84]

    , author Sreedharan, S

    author Valmeekam, K. , author Sreedharan, S. , author Marquez, M. , author Olmo, A. , author Kambhampati, S. , year 2023 b. title On the planning abilities of large language models (a critical investigation with a proposed benchmark) . https://arxiv.org/abs/2302.06706, http://...

    2023 arXiv

  77. [85]

    , author Shazeer, N

    author Vaswani, A. , author Shazeer, N. , author Parmar, N. , author Uszkoreit, J. , author Jones, L. , author Gomez, A.N. , author Kaiser, L.u. , author Polosukhin, I. , year 2017 . title Attention is all you need . journal Advances in Neural Information Processing Systems vo...

    2017

  78. [86]

    , author Bonatti, R

    author Vemprala, S. , author Bonatti, R. , author Bucker, A. , author Kapoor, A. , year 2023 . title ChatGPT for Robotics: Design Principles and Model Abilities . type Technical Report number MSR-TR-2023-8 . Microsoft. https://www.microsoft.com/en-us/research/publication/chatg...

    2023

  79. [87]

    , author Li, X

    author Wang, J. , author Li, X. , author Li, J. , author Sun, Q. , author Wang, H. , year 2022 . title Ngcu: A new rnn model for time-series data prediction . journal Big Data Research volume 27 , pages 100296

    2022

  80. [88]

    , author Wang, X

    author Wei, J. , author Wang, X. , author Schuurmans, D. , author Bosma, M. , author Ichter, B. , author Xia, F. , author Chi, E. , author Le, Q. , author Zhou, D. , year 2023 . title Chain-of-thought prompting elicits reasoning in large language models . http://arxiv.org/abs/...

    2023 arXiv

  81. [89]

    , author Reichert, M

    author Wewerka, J. , author Reichert, M. , year 2023 . title Robotic process automation-a systematic mapping study and classification framework . journal Enterprise Information Systems volume 17 , pages 1986862

    2023

  82. [90]

    Baltija Publishing

    author Yakovenko, Y. , author Shaptala, R. , year 2023 . title Intelligent process automation, robotic process automation and artificial intelligence for business processes transformation . journal Publishing House “Baltija Publishing”

    2023

  83. [91]

    , author Wang, X

    author Yang, C. , author Wang, X. , author Lu, Y. , author Liu, H. , author Le, Q.V. , author Zhou, D. , author Chen, X. , year 2024 a. title Large language models as optimizers . journal The Twelfth International Conference on Learning Representations https://openreview.net/f...

    2024

  84. [92]

    , author Ning, L

    author Yang, Z. , author Ning, L. , author Wang, H. , author Jiang, T. , author Zhang, S. , author Cui, S. , author Jiang, H. , author Li, C. , author Wang, S. , author Wang, Z. , year 2024 b. title Text2reaction : Enabling reactive task planning using large language models . ...

    2024

  85. [93]

    , author Teng, J

    author Yang, Z. , author Teng, J. , author Zheng, W. , author Ding, M. , author Huang, S. , author Xu, J. , author Yang, Y. , author Hong, W. , author Zhang, X. , author Feng, G. , et al., year 2024 c. title Cogvideox: Text-to-video diffusion models with an expert transformer ...

    2024 arXiv

  86. [94]

    , author Kim, I.Y

    author Yu, W. , author Kim, I.Y. , author Mechefske, C. , year 2021 . title Analysis of different rnn autoencoder variants for time series classification and machine prognostics . journal Mechanical Systems and Signal Processing volume 149 , pages 107322

    2021

  87. [95]

    , author Watson, W

    author Zeng, Z. , author Watson, W. , author Cho, N. , author Rahimi, S. , author Reynolds, S. , author Balch, T. , author Veloso, M. , year 2024 . title Flowmind: Automatic workflow generation with llms . https://arxiv.org/abs/2404.13050, http://arxiv.org/abs/2404.13050 arXiv...

    2024 arXiv

  88. [96]

    , author Wallace, E

    author Zhao, T.Z. , author Wallace, E. , author Feng, S. , author Klein, D. , author Singh, S. , year 2021 . title Calibrate before use: Improving few-shot performance of language models . https://arxiv.org/abs/2102.09690, http://arxiv.org/abs/2102.09690 arXiv:2102.09690

    2021 arXiv

  89. [97]

    , author Song, J

    author Zhou, Z. , author Song, J. , author Yao, K. , author Shu, Z. , author Ma, L. , year 2024 . title Isr-llm: Iterative self-refined large language model for long-horizon sequential task planning , in: booktitle 2024 IEEE International Conference on Robotics and Automation ...

    2024

Pith tools

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