Pith. sign in

REVIEW 3 major objections 4 minor 90 references

Who is a Better Player: LLM against LLM

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Board-game tournaments for LLMs show optimism under pressure and unstable skill.

desk verdict The submission is unverifiable as an LLM benchmark paper because the full text is an unrelated SEM reconstruction paper; the abstract alone is not enough to evaluate the claims. read the letter →

arxiv 2508.04720 v1 pith:GGTC2QAA submitted 2025-08-05 cs.AI

Yingjie Zhou , Jiezhang Cao , Farong Wen , Li Xu , Yanwei Jiang , Jun Jia , Ronghui Li , Xiaohong Liu
show 5 more authors
Yu Zhou Xiongkuo Min Jie Guo Zicheng Zhang Guangtao Zhai
This is my paper
classification cs.AI
keywords LLMevaluationadversarialboardgamesEloratingPerformanceLoopGraphsentimentanalysisround-robintournamentQiTown
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 competitive board games can serve as a more revealing benchmark for large language models than question-and-answer tests, because they exercise strategic reasoning, adaptability, and psychological responses under adversarial pressure. It introduces Qi Town, a platform where 20 LLM-driven players compete in five games, and evaluates them with Elo ratings, a Performance Loop Graph, and a Positive Sentiment Score computed from in-game text. The reported results claim that most LLMs keep a positive outlook whether winning or losing, suggesting greater stress resilience than human players, while cyclic win-loss patterns in the graph expose instability in their skill. A sympathetic reader would care because it proposes a way to assess qualities in AI that conventional benchmarks ignore.

What carries the argument

Qi Town is the proposed evaluation platform: it runs round-robin tournaments across five games among twenty LLM-driven players. Ratings come from the Elo system, the Performance Loop Graph (PLG) visualizes non-transitive win-loss cycles in the tournament, and the Positive Sentiment Score (PSS) is computed from the language the models produce during the game as a proxy for mental fitness. The PLG and PSS are the paper's new instruments, and the claims rest on what they measure.

What would settle it

Run the same round-robin tournament with human or random-move players; if comparable win-loss cycles appear, the PLG cycles are not evidence of LLM instability, and score sentiment of LLMs against a matched human control group under identical game pressure to test the claimed greater adaptability.

Watch

Extended reading notes

Core claim

The central claim is that, despite differences in underlying models, LLM agents generally maintain optimism through winning and losing streaks, a trait the authors interpret as greater adaptability to high-stress adversarial situations than humans exhibit. Alongside this, the Performance Loop Graph, which maps cyclic win-loss relations among players, reveals that the skill level of most LLMs is unstable during play because their outcomes loop rather than forming a consistent hierarchy. If true, this means sentiment and stability of play are measurable properties of LLM behavior, and board-game tournaments can expose them.

Load-bearing premise

The claim that cyclic win-loss patterns reveal unstable skill assumes that such cycles are not a natural property of the games themselves, and the supplied full text does not match the abstract, so the experimental basis for these claims is not present in this manuscript.

Editorial extensions

If this is right

  • If LLMs stay optimistic under competitive stress, sentiment-based metrics could join accuracy scores in LLM evaluation.
  • If the PLG's cycles reflect genuine instability, rankings from single matches are unreliable and longer tournaments are needed for fair comparison.
  • Adversarial board games could become a standard complement to Q&A benchmarks for judging reasoning and adaptability.
  • PSS might transfer to real-world AI applications where sustained optimism under failure matters, such as negotiation or customer service.
  • The framework suggests that comparing LLM personalities under stress is feasible using text-only signals.

Reading between the lines

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

  • One inference is that the interpretation of cycles as instability assumes non-transitive outcomes are not a natural feature of the games; if those games have intrinsic rock-paper-scissors dynamics, cycles would appear even for perfectly stable players.
  • Another inference is that the positive-sentiment result could be an artifact of LLM training to be agreeable or polite rather than a genuine psychological trait, so a test would be to check whether sentiment tracks the actual game state or only the surface text.
  • The comparative claim of being more adaptable than humans requires a matched human control group under identical game pressure; absent that, it is a suggestive observation, not a demonstrated difference.
  • The supplied full text is a different manuscript on SEM 3D reconstruction, so the abstract's experimental claims currently lack any described evidence in this document; a reader cannot verify the reported tournament results from what is given.
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

3 major / 4 minor

Summary. The manuscript, as submitted, consists of an abstract proposing an adversarial board-game benchmarking framework for LLMs, and a full text that is an entirely different paper on neural-field 3D surface reconstruction from SEM images. The abstract introduces the Qi Town platform, five games, twenty LLM-driven players, Elo ratings, a Performance Loop Graph (PLG), and a Positive Sentiment Score (PSS), and claims experimental findings that most LLMs remain optimistic about winning and losing, show greater adaptability to high-stress adversarial environments than humans, and exhibit cyclic win-loss patterns that expose instability of skill play. The full text, however, is the NFH-SEM paper (arXiv:2508.04728v2), with different authors, title, and content, containing no mention of Qi Town, LLM players, Elo, PLG, PSS, or any board-game experiment. Consequently, the submitted record contains no experimental design, data, metric definitions, or analysis that could support or refute the abstract's central claims.

Significance. If the claimed framework and results were properly documented, the paper could offer a useful complement to Q&A-based LLM benchmarks by evaluating strategic reasoning through board-game competitions and by attempting to measure affective states (PSS) during play. The PLG idea for visualizing cyclic win-loss relationships is also potentially interesting. However, as submitted, the manuscript provides no support for these contributions: there are no definitions of the proposed metrics, no description of the tournament protocol, no data, no statistical analysis, and no comparison to human play. There are also no machine-checked proofs, reproducible code, or falsifiable predictions that could be incrementally verified. The potential significance is therefore entirely prospective and cannot be assessed from the present record.

major comments (3)
  1. [Full text (header and entirety)] The submitted full text is a different manuscript: 'Neural Field-Based 3D Surface Reconstruction of Microstructures from Multi-Detector Signals in Scanning Electron Microscopy', carrying the arXiv identifier 2508.04728v2 and a different author list. None of the abstract's load-bearing elements—Qi Town, the five games, the twenty LLM players, the round-robin tournament, Elo ratings, PLG, or PSS—appear anywhere in the body. The manuscript therefore contains no experimental design, no data, no metric definitions, and no statistical analysis corresponding to the abstract's claims. This is not a local omission but a complete absence of the subject matter, so the central claims are unevaluable and the error cannot be fixed by revising individual sections.
  2. [Abstract (PLG interpretation)] The abstract's inference from 'the complex relationship between cyclic wins and losses in PLGs' to 'the instability of LLMs' skill play' is not logically forced: non-transitive win-loss cycles can arise from the intrinsic strategic structure of the games themselves (e.g., rock-paper-scissors-like dominance relations) without any within-player performance instability. The abstract provides no definition of PLG, no null model for cycle prevalence, and no control for game-dependent non-transitivity, so the stated conclusion would not follow even from the claimed experimental results.
  3. [Abstract (human comparison)] The claim that 'most LLMs ... demonstrating greater adaptability to high-stress adversarial environments than humans' requires a human baseline, a definition of adaptability, and a specification of how 'high-stress' was induced or measured. None of these is provided in the abstract or the full text, and no human-subject protocol or human performance data are described. This comparative claim is therefore unsupported on the submitted record.
minor comments (4)
  1. [Abstract (experimental reporting)] The abstract reports 'experimental results' without any indication of sample sizes, number of games per pairing, error bars, or statistical tests; even setting aside the full-text mismatch, this level of reporting would be insufficient for the claimed quantitative comparisons.
  2. [Abstract (metric definitions)] PSS and PLG are introduced by name but never defined; a complete submission should provide formal definitions, the data used to compute them, and the procedure for aggregating game outcomes into the PLG.
  3. [References] The abstract cites no prior work on game-based LLM evaluation, adversarial benchmarks, or the use of Elo ratings in AI systems; the manuscript would benefit from situating the proposed framework relative to existing benchmarks.
  4. [Submission consistency] The body text's header arXiv:2508.04728v2 does not match the submitted arXiv:2508.04720; the authors should correct the submission so that the abstract and full text describe the same work.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identified: the abstract makes empirical claims with no derivation chain, and the supplied full text is an unrelated manuscript, making the claims unevaluable rather than circular.

full rationale

The claimed derivation chain in the abstract is entirely empirical: a round-robin board-game tournament among LLMs, Elo ratings, a Performance Loop Graph, and a Positive Sentiment Score are asserted as experimental instruments, and the conclusions are stated as observed results. No equation, fitted parameter, or metric definition appears in the abstract, so there is no derivation that could reduce to its own inputs. The supplied full text is a different paper on neural-field 3D reconstruction from SEM images (NFH-SEM, arXiv:2508.04728v2), with different authors and no mention of Qi Town, the five games, the 20 LLM players, Elo, PLG, PSS, or the tournament. This mismatch means the abstract's claims cannot be checked at all from the submitted record, but an absent derivation or missing experimental description is not circularity under the stated rules: no specific reduction, no fitted parameter renamed as a prediction, no load-bearing self-citation, and no uniqueness claim imported from prior work can be quoted. The interpretive weakness noted in the skeptical reading, namely that cyclic win-loss patterns may reflect game structure rather than skill instability, is a validity concern about the inference from data to conclusion, not a circularity of the derivation itself. Therefore the appropriate finding is no significant circularity, with score 0, while noting that the paper cannot be substantively evaluated due to the full-text mismatch.

Assumptions & free parameters 0 free parameters · 2 assumptions · 1 invented entities

Because the attached full text is a different paper, the ledger is reconstructed from the abstract alone. No free parameters can be audited, and the two axioms listed are interpretive assumptions the abstract makes without support. Qi Town is the only named artifact, and it has no independent evidence.

assumptions (2)
  • domain assumption Board-game performance is a valid proxy for comprehensive LLM capability
    The abstract motivates the benchmark by claiming it compensates Q&A data dependence, implying that strategic game play measures LLM intelligence broadly. This is asserted, not argued.
  • ad hoc to paper Cyclic win-loss relationships in the Performance Loop Graph indicate instability of skill play
    The abstract interprets cyclic wins and losses as 'instability of LLMs' skill play', assuming that non-transitive outcomes reflect performance variability rather than game structure. No evidence is given.
invented entities (1)
  • Qi Town
    purpose: Specialized evaluation platform hosting LLM-driven players in board games
    The abstract names Qi Town as a new platform, but no code, dataset, or documentation appears in the body text. No external falsifiable handle is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Who is a Better Player: LLM against LLM." pith.science (2026). https://pith.science/paper/GGTC2QAA

@misc{pith2026250804720,
  author       = {Pith},
  title        = {Pith review of: Who is a Better Player: LLM against LLM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GGTC2QAA}},
  note         = {Machine review of arXiv:2508.04720}
}
read the original abstract

Adversarial board games, as a paradigmatic domain of strategic reasoning and intelligence, have long served as both a popular competitive activity and a benchmark for evaluating artificial intelligence (AI) systems. Building on this foundation, we propose an adversarial benchmarking framework to assess the comprehensive performance of Large Language Models (LLMs) through board games competition, compensating the limitation of data dependency of the mainstream Question-and-Answer (Q&A) based benchmark method. We introduce Qi Town, a specialized evaluation platform that supports 5 widely played games and involves 20 LLM-driven players. The platform employs both the Elo rating system and a novel Performance Loop Graph (PLG) to quantitatively evaluate the technical capabilities of LLMs, while also capturing Positive Sentiment Score (PSS) throughout gameplay to assess mental fitness. The evaluation is structured as a round-robin tournament, enabling systematic comparison across players. Experimental results indicate that, despite technical differences, most LLMs remain optimistic about winning and losing, demonstrating greater adaptability to high-stress adversarial environments than humans. On the other hand, the complex relationship between cyclic wins and losses in PLGs exposes the instability of LLMs' skill play during games, warranting further explanation and exploration.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

90 extracted references · 74 canonical work pages

  1. [1]

    3D SEM metrology of microstructures for high volume manu- facturing

    Zeinab Abdallah, Aur ´elien Fay, and St ´ephane Bonnet. 3D SEM metrology of microstructures for high volume manu- facturing. In38th European Mask and Lithography Confer- ence (EMLC 2023), pages 229–248. SPIE, 2023. 2

    2023

  2. [2]

    A survey of photo- metric stereo techniques.Found

    Jens Ackermann and Michael Goesele. A survey of photo- metric stereo techniques.Found. Trends. Comput. Graph. Vis., 9(3–4):149–254, 2015. 2

    2015

  3. [3]

    Agisoft Metashape

    Agisoft LLC. Agisoft Metashape. Computer software, 2025. 4, 5, 6

    2025

  4. [4]

    A. D. Ball, P. A. Job, and A. E. L. Walker. SEM- microphotogrammetry, a new take on an old method for gen- erating high-resolution 3D models from SEM images.Jour- nal of microscopy, 267(2):214–226, 2017. 2

    2017

  5. [5]

    Investigation of the effect of magnification, acceler- ating voltage, and working distance on the 3D digital recon- struction techniques.Scanning, 2020(1):3743267, 2020

    Seyed Mahmoud Bayazid, Nicolas Brodusch, and Raynald Gauvin. Investigation of the effect of magnification, acceler- ating voltage, and working distance on the 3D digital recon- struction techniques.Scanning, 2020(1):3743267, 2020. 2, 4

    2020

  6. [6]

    shape from shading

    Wolfgang Beil and IC Carlsen. Surface reconstruction from stereoscopy and “shape from shading” in SEM images.Ma- chine vision and applications, 4(4):271–285, 1991. 2

    1991

  7. [7]

    Interpretation of hydrogen-assisted fatigue crack propagation in BCC iron based on dislocation structure evo- lution around the crack wake.Acta Materialia, 156:245–253,

    Domas Birenis, Yuhei Ogawa, Hisao Matsunaga, Osamu Takakuwa, Junichiro Yamabe, Oystein Prytz, and Annett Thogersen. Interpretation of hydrogen-assisted fatigue crack propagation in BCC iron based on dislocation structure evo- lution around the crack wake.Acta Materialia, 156:245–253,

  8. [8]

    MVSFormer++: Revealing the devil in transformer’s details for multi-view stereo

    Chenjie Cao, Xinlin Ren, and Yanwei Fu. MVSFormer++: Revealing the devil in transformer’s details for multi-view stereo. InInternational Conference on Learning Represen- tations (ICLR), 2024. 15, 16, 17

    2024

Show all 90 references
  1. [9]

    PGSR: Planar-based gaussian splat- ting for efficient and high-fidelity surface reconstruction

    Danpeng Chen, Hai Li, Weicai Ye, Yifan Wang, Weijian Xie, Shangjin Zhai, Nan Wang, Haomin Liu, Hujun Bao, and Guofeng Zhang. PGSR: Planar-based gaussian splat- ting for efficient and high-fidelity surface reconstruction. IEEE Transactions on Visualization and Computer Graph- i...

    2024

  2. [10]

    Multi-view neural 3D reconstruction of micro-and nanostructures with atomic force microscopy.Communications Engineering, 3(1):131,

    Shuo Chen, Mao Peng, Yijin Li, Bing-Feng Ju, Hujun Bao, Yuan-Liu Chen, and Guofeng Zhang. Multi-view neural 3D reconstruction of micro-and nanostructures with atomic force microscopy.Communications Engineering, 3(1):131,

  3. [11]

    OpticFu- sion: Multi-modal neural implicit 3D reconstruction of mi- crostructures by fusing white light interferometry and optical microscopy

    Shuo Chen, Yijin Li, and Guofeng Zhang. OpticFu- sion: Multi-modal neural implicit 3D reconstruction of mi- crostructures by fusing white light interferometry and optical microscopy. In2025 International Conference on 3D Vision (3DV), pages 90–100, 2025. 3

    2025

  4. [12]

    Integrating shape from shading and shape from stereo for variable reflectance sur- face reconstruction from SEM images

    Reinhard Danzl and Stefan Scherer. Integrating shape from shading and shape from stereo for variable reflectance sur- face reconstruction from SEM images. InProceedings 26th Workshop of the Austrian Association for Pattern Recogni- tion (AAPR), pages 281–288, 2002. 2

    2002

  5. [13]

    TransMVS- Net: Global context-aware multi-view stereo network with transformers

    Yikang Ding, Wentao Yuan, Qingtian Zhu, Haotian Zhang, Xiangyue Liu, Yuanjiang Wang, and Xiao Liu. TransMVS- Net: Global context-aware multi-view stereo network with transformers. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8585– ...

    2022

  6. [14]

    Three-dimensional characterization of microstruc- tures in a SEM.Measurement Science and Technology, 17 (1):28, 2005

    Wlodzimierz Drzazga, Jaroslaw Paluszynski, and Witold Sł´owko. Three-dimensional characterization of microstruc- tures in a SEM.Measurement Science and Technology, 17 (1):28, 2005. 2

    2005

  7. [15]

    Confocal microscopy: principles and mod- ern practices.Current protocols in cytometry, 92(1):e68,

    Amicia D Elliott. Confocal microscopy: principles and mod- ern practices.Current protocols in cytometry, 92(1):e68,

  8. [16]

    3D reconstruction of SEM images by use of optical photogrammetry software.Journal of structural biology, 191(2):190–196, 2015

    Mona Eulitz and Gebhard Reiss. 3D reconstruction of SEM images by use of optical photogrammetry software.Journal of structural biology, 191(2):190–196, 2015. 1, 2, 3, 4, 6, 7

    2015

  9. [17]

    Investigations on the pollen morphology of some fruit species.Turkish Journal of Agriculture and Forestry, 33(2):181–190, 2009

    Yasemin Evrenoso ˘glu and Adalet Misirli. Investigations on the pollen morphology of some fruit species.Turkish Journal of Agriculture and Forestry, 33(2):181–190, 2009. 6

    2009

  10. [18]

    Transformation of hard pollen into soft matter.Nature Communications, 11(1):1449, 2020

    Teng-Fei Fan, Soohyun Park, Qian Shi, Xingyu Zhang, Qimin Liu, Yoohyun Song, Hokyun Chin, Mo- hammed Shahrudin Bin Ibrahim, Natalia Mokrzecka, Yun Yang, Hua Li, Juha Song, Subra Suresh, and Nam-Joon Cho. Transformation of hard pollen into soft matter.Nature Communications, 11(...

    2020

  11. [19]

    Regularization techniques for 3D surface reconstruction from four quadrant backscattered electron detector images.Ultramicroscopy, 250:113746,

    Matteo Giardino, Devanarayanan Meena Narayana Menon, and Davide Luca Janner. Regularization techniques for 3D surface reconstruction from four quadrant backscattered electron detector images.Ultramicroscopy, 250:113746,

  12. [20]

    Golek, P

    F. Golek, P. Mazur, Z. Ryszka, and S. Zuber. AFM image artifacts.Applied Surface Science, 304:11–19, 2014. 14

    2014

  13. [21]

    Implicit geometric regularization for learning shapes

    Amos Gropp, Lior Yariv, Niv Haim, Matan Atzmon, and Yaron Lipman. Implicit geometric regularization for learning shapes. InProceedings of the 37th International Conference on Machine Learning, pages 3789–3799, 2020. 4, 5

    2020

  14. [22]

    Cascade cost volume for high-resolution multi-view stereo and stereo matching

    Xiaodong Gu, Zhiwen Fan, Siyu Zhu, Zuozhuo Dai, Feitong Tan, and Ping Tan. Cascade cost volume for high-resolution multi-view stereo and stereo matching. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2495–2504, 2020. 2

    2020

  15. [23]

    SuGaR: Surface- aligned gaussian splatting for efficient 3D mesh reconstruc- tion and high-quality mesh rendering

    Antoine Gu ´edon and Vincent Lepetit. SuGaR: Surface- aligned gaussian splatting for efficient 3D mesh reconstruc- tion and high-quality mesh rendering. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5354–5363, 2024. 3

    2024

  16. [24]

    Instant neural surface reconstruction, 2022

    Yuan-Chen Guo. Instant neural surface reconstruction, 2022. https://github.com/bennyguo/instant-nsr-pl. 13 9

    2022

  17. [25]

    Two-photon lithography for three-dimensional fabrication in micro/nanoscale regime: A comprehensive review.Optics & Laser Technology, 142: 107180, 2021

    V Harinarayana and YC Shin. Two-photon lithography for three-dimensional fabrication in micro/nanoscale regime: A comprehensive review.Optics & Laser Technology, 142: 107180, 2021. 6

    2021

  18. [26]

    Least squares surface re- construction from measured gradient fields

    Matthew Harker and Paul O’Leary. Least squares surface re- construction from measured gradient fields. In2008 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 1–7, 2008. 6

    2008

  19. [27]

    Matthew Harker and Paul O’Leary. Regularized reconstruc- tion of a surface from its measured gradient field algorithms for spectral, tikhonov, constrained, and weighted regulariza- tion.Journal of Mathematical Imaging and Vision, 51(1): 46–70, 2015

    2015

  20. [28]

    Direct regularized sur- face reconstruction from gradients for industrial photometric stereo.Computers in industry, 64(9):1221–1228, 2013

    Matthew Harker and Paul O’Leary. Direct regularized sur- face reconstruction from gradients for industrial photometric stereo.Computers in industry, 64(9):1221–1228, 2013. 6

    2013

  21. [29]

    2D gaussian splatting for geometrically ac- curate radiance fields

    Binbin Huang, Zehao Yu, Anpei Chen, Andreas Geiger, and Shenghua Gao. 2D gaussian splatting for geometrically ac- curate radiance fields. InACM SIGGRAPH 2024 conference papers, pages 1–11, 2024. 2, 3, 7

    2024

  22. [30]

    Structure from motion photogrammetry in forestry: A review.Current Forestry Reports, 5(3):155–168, 2019

    Jakob Iglhaut, Carlos Cabo, Stefano Puliti, Livia Piermat- tei, James O’Connor, and Jacqueline Rosette. Structure from motion photogrammetry in forestry: A review.Current Forestry Reports, 5(3):155–168, 2019. 2

    2019

  23. [31]

    Large scale multi-view stereopsis eval- uation

    Rasmus Jensen, Anders Dahl, George V ogiatzis, Engin Tola, and Henrik Aanæs. Large scale multi-view stereopsis eval- uation. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 406–413, 2014. 15

    2014

  24. [32]

    Coordinate-based neural representations for computational adaptive optics in widefield microscopy.Nature Machine Intelligence, 6(6):714–725, 2024

    Iksung Kang, Qinrong Zhang, Stella X Yu, and Na Ji. Coordinate-based neural representations for computational adaptive optics in widefield microscopy.Nature Machine Intelligence, 6(6):714–725, 2024. 3

    2024

  25. [33]

    Finer features for functional microdevices - micromachines can be created with higher resolution using two-photon absorption.Nature, 412(6848):697–698, 2001

    S Kawata, HB Sun, T Tanaka, and K Takada. Finer features for functional microdevices - micromachines can be created with higher resolution using two-photon absorption.Nature, 412(6848):697–698, 2001. 6

    2001

  26. [34]

    MapAnything: Universal feed-forward metric 3D re- construction.arXiv preprint arXiv:2509.13414, 2025

    Nikhil Keetha, Norman M ¨uller, Johannes Sch ¨onberger, Lorenzo Porzi, Yuchen Zhang, Tobias Fischer, Arno Knapitsch, Duncan Zauss, Ethan Weber, Nelson Antunes, et al. MapAnything: Universal feed-forward metric 3D re- construction.arXiv preprint arXiv:2509.13414, 2025. 2, 7

    2025 arXiv

  27. [35]

    3D Gaussian splatting for real-time radiance field rendering.ACM Trans

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3D Gaussian splatting for real-time radiance field rendering.ACM Trans. Graph., 42(4):139–1,

  28. [36]

    Adam: A method for stochastic opti- mization.arXiv preprint arXiv:1412.6980, 2014

    Diederik P Kingma. Adam: A method for stochastic opti- mization.arXiv preprint arXiv:1412.6980, 2014. 5

    2014 arXiv

  29. [37]

    Extracting three-dimensional informa- tion from SEM images by means of photogrammetry.Mi- cron, 134:102873, 2020

    Paweł Kozikowski. Extracting three-dimensional informa- tion from SEM images by means of photogrammetry.Mi- cron, 134:102873, 2020. 1, 2, 3, 4, 6, 7

    2020

  30. [38]

    Fundamental aspects of resolution and precision in vertical scanning white-light interferometry.Surface Topography: Metrology and Properties, 4(2):024004, 2016

    Peter Lehmann, Stanislav Tereschenko, and Weichang Xie. Fundamental aspects of resolution and precision in vertical scanning white-light interferometry.Surface Topography: Metrology and Properties, 4(2):024004, 2016. 14

    2016

  31. [39]

    Ground- ing image matching in 3D with MASt3R

    Vincent Leroy, Yohann Cabon, and J´erˆome Revaud. Ground- ing image matching in 3D with MASt3R. InEuropean Con- ference on Computer Vision, pages 71–91. Springer, 2024. 2

    2024

  32. [40]

    Neuralangelo: High-fidelity neural surface reconstruction

    Zhaoshuo Li, Thomas M ¨uller, Alex Evans, Russell H Tay- lor, Mathias Unberath, Ming-Yu Liu, and Chen-Hsuan Lin. Neuralangelo: High-fidelity neural surface reconstruction. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 8456–8465, 2023. 3

    2023

  33. [41]

    Carson Meredith

    Haisheng Lin, Ismael Gomez, and J. Carson Meredith. Pol- lenkitt wetting mechanism enables species-specific tunable pollen adhesion.Langmuir, 29(9):3012–3023, 2013. 1, 6

    2013

  34. [42]

    Recovery of continuous 3D refractive index maps from discrete intensity-only measurements using neu- ral fields.Nature Machine Intelligence, 4(9):781–791, 2022

    Renhao Liu, Yu Sun, Jiabei Zhu, Lei Tian, and Ulug- bek S Kamilov. Recovery of continuous 3D refractive index maps from discrete intensity-only measurements using neu- ral fields.Nature Machine Intelligence, 4(9):781–791, 2022. 3

    2022

  35. [43]

    Marching cubes: A high resolution 3D surface construction algorithm

    William E Lorensen and Harvey E Cline. Marching cubes: A high resolution 3D surface construction algorithm. InSemi- nal graphics: pioneering efforts that shaped the field, pages 347–353. 1998. 5, 13

    1998

  36. [44]

    NeRF: Representing scenes as neural radiance fields for view syn- thesis.Communications of the ACM, 65(1):99–106, 2021

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. NeRF: Representing scenes as neural radiance fields for view syn- thesis.Communications of the ACM, 65(1):99–106, 2021. 3

    2021

  37. [45]

    Instant neural graphics primitives with a mul- tiresolution hash encoding.ACM Transactions on Graphics (ToG), 41(4):1–15, 2022

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. Instant neural graphics primitives with a mul- tiresolution hash encoding.ACM Transactions on Graphics (ToG), 41(4):1–15, 2022. 4, 13

    2022

  38. [46]

    Fatigue behavior of an ultrafine-grained Al-Mg-Si alloy processed by high-pressure torsion.Metals, 5(2):578– 590, 2015

    Maxim Murashkin, Ilchat Sabirov, Dmitriy Prosvirnin, Ilya Ovid’ko, Vladimir Terentiev, Ruslan Valiev, and Sergey Do- batkin. Fatigue behavior of an ultrafine-grained Al-Mg-Si alloy processed by high-pressure torsion.Metals, 5(2):578– 590, 2015. 6

    2015

  39. [47]

    Principal image decomposi- tion for multi-detector backscatter electron topography re- construction.Ultramicroscopy, 227:113200, 2021

    Jan Neggers, Eva H ´eripr´e, Marc Bonnet, Denis Boivin, Alexandre Tanguy, Simon Hallais, Fabrice Gaslain, Elodie Rouesne, and St ´ephane Roux. Principal image decomposi- tion for multi-detector backscatter electron topography re- construction.Ultramicroscopy, 227:113200, 2021. 2, 16

    2021

  40. [48]

    Nelson, Neville L

    Erik C. Nelson, Neville L. Dias, Kevin P. Bassett, Simon N. Dunham, Varun Verma, Masao Miyake, Pierre Wiltzius, John A. Rogers, James J. Coleman, Xiuling Li, and Paul V . Braun. Epitaxial growth of three-dimensionally architec- tured optoelectronic devices.Nature Materials, 10...

    2011

  41. [49]

    UNISURF: Unifying neural implicit surfaces and radiance fields for multi-view reconstruction

    Michael Oechsle, Songyou Peng, and Andreas Geiger. UNISURF: Unifying neural implicit surfaces and radiance fields for multi-view reconstruction. InProceedings of the IEEE/CVF international conference on computer vision, pages 5589–5599, 2021. 2, 3

    2021

  42. [50]

    A survey of structure from motion.Acta Numerica, 26:305–364, 2017

    Onur ¨Ozyes ¸il, Vladislav V oroninski, Ronen Basri, and Amit Singer. A survey of structure from motion.Acta Numerica, 26:305–364, 2017. 2

    2017

  43. [51]

    Surface reconstruction with the photometric method in SEM.Vacuum, 78(2-4):533–537,

    J Paluszy ´nski and W Sł ´owko. Surface reconstruction with the photometric method in SEM.Vacuum, 78(2-4):533–537,

  44. [52]

    PyTorch: An imperative style, high-performance deep learning li- 10 brary.Advances in neural information processing systems, 32:8024–8035, 2019

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zem- ing Lin, Natalia Gimelshein, Luca Antiga, et al. PyTorch: An imperative style, high-performance deep learning li- 10 brary.Advances in neural information processing system...

    2019

  45. [53]

    Rethinking depth estimation for multi- view stereo: A unified representation

    Rui Peng, Rongjie Wang, Zhenyu Wang, Yawen Lai, and Ronggang Wang. Rethinking depth estimation for multi- view stereo: A unified representation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8645–8654, 2022. 2

    2022

  46. [54]

    Adam L. Pilchak. Fatigue crack growth rates in alpha tita- nium: Faceted vs. striation growth.Scripta Materialia, 68 (5):277–280, 2013. 6

    2013

  47. [55]

    An automatic alignment procedure for a four-source photomet- ric stereo technique applied to scanning electron microscopy

    Ruggero Pintus, Simona Podda, and Massimo Vanzi. An automatic alignment procedure for a four-source photomet- ric stereo technique applied to scanning electron microscopy. IEEE Transactions on Instrumentation and Measurement, 57 (5):989–996, 2008. 2, 4

    2008

  48. [56]

    The effect of spatial micro-CT image resolution and surface complexity on the morphological 3D analysis of open porous structures.Materials Characterization, 87:104– 115, 2014

    Grzegorz Pyka, Greet Kerckhofs, Jan Schrooten, and Mar- tine Wevers. The effect of spatial micro-CT image resolution and surface complexity on the morphological 3D analysis of open porous structures.Materials Characterization, 87:104– 115, 2014. 14

    2014

  49. [57]

    Springer, Berlin / Hei- delberg, 1998

    Ludwig Reimer.Scanning Electron Microscopy: Physics of Image Formation and Microanalysis. Springer, Berlin / Hei- delberg, 1998. 1, 3, 4, 8

    1998

  50. [58]

    Sch ¨onberger and Jan-Michael Frahm

    Johannes L. Sch ¨onberger and Jan-Michael Frahm. Structure- from-motion revisited. In2016 IEEE Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 4104– 4113, 2016. 2

    2016

  51. [59]

    A benchmark dataset and evalua- tion for non-lambertian and uncalibrated photometric stereo

    Boxin Shi, Zhipeng Mo, Zhe Wu, Dinglong Duan, Sai-Kit Yeung, and Ping Tan. A benchmark dataset and evalua- tion for non-lambertian and uncalibrated photometric stereo. IEEE Transactions on Pattern Analysis and Machine Intelli- gence, 41(2):271–284, 2019. 2

    2019

  52. [60]

    Specific features of the miniature ionisa- tion BSE multi-detector unit for 3D imaging in environmen- tal conditions.Micron, 126:102752, 2019

    Witold Sł ´owko. Specific features of the miniature ionisa- tion BSE multi-detector unit for 3D imaging in environmen- tal conditions.Micron, 126:102752, 2019. 2, 3, 4, 13, 16

    2019

  53. [61]

    Detector system for three- dimensional imaging in the variable pressure/environmental SEM.Acta Physica Polonica A, 123(5):877–879, 2013

    W Sł ´owko and M Krysztof. Detector system for three- dimensional imaging in the variable pressure/environmental SEM.Acta Physica Polonica A, 123(5):877–879, 2013. 2, 3, 4, 6, 16

    2013

  54. [62]

    Real three-dimensional microstruc- tures fabricated by photopolymerization of resins through two-photon absorption.Optics letters, 25(15):1110–1112,

    HB Sun, T Kawakami, Y Xu, JY Ye, S Matuso, H Misawa, M Miwa, and R Kaneko. Real three-dimensional microstruc- tures fabricated by photopolymerization of resins through two-photon absorption.Optics letters, 25(15):1110–1112,

  55. [63]

    Generating open-source 3D phytoplankton models by integrating photogrammetry with scanning electron microscopy.Frontiers in Microbi- ology, 15:1429179, 2024

    Xuerong Sun, Robert JW Brewin, Christian Hacker, Jo- hannes J Viljoen, and Mengyu Li. Generating open-source 3D phytoplankton models by integrating photogrammetry with scanning electron microscopy.Frontiers in Microbi- ology, 15:1429179, 2024. 2, 6

    2024

  56. [64]

    Tafti, Andrew B

    Ahmad P. Tafti, Andrew B. Kirkpatrick, Zahrasadat Alavi, Heather A. Owen, and Zeyun Yu. Recent advances in 3D SEM surface reconstruction.Micron, 78:54–66, 2015. 1, 2, 3, 7

    2015

  57. [65]

    BA-Net: Dense bundle ad- justment network.arXiv preprint arXiv:1806.04807, 2018

    Chengzhou Tang and Ping Tan. BA-Net: Dense bundle ad- justment network.arXiv preprint arXiv:1806.04807, 2018. 2

    2018 arXiv

  58. [66]

    DROID-SLAM: Deep visual SLAM for monocular, stereo, and RGB-D cameras.Ad- vances in neural information processing systems, 34:16558– 16569, 2021

    Zachary Teed and Jia Deng. DROID-SLAM: Deep visual SLAM for monocular, stereo, and RGB-D cameras.Ad- vances in neural information processing systems, 34:16558– 16569, 2021

    2021

  59. [67]

    Deep patch vi- sual odometry.Advances in Neural Information Processing Systems, 36:39033–39051, 2023

    Zachary Teed, Lahav Lipson, and Jia Deng. Deep patch vi- sual odometry.Advances in Neural Information Processing Systems, 36:39033–39051, 2023. 2

    2023

  60. [68]

    DN-Splatter: Depth and normal priors for gaussian splatting and meshing

    Matias Turkulainen, Xuqian Ren, Iaroslav Melekhov, Otto Seiskari, Esa Rahtu, and Juho Kannala. DN-Splatter: Depth and normal priors for gaussian splatting and meshing. In 2025 IEEE/CVF Winter Conference on Applications of Com- puter Vision (WACV), pages 2421–2431. IEEE, 2025. 3, 7

    2025

  61. [69]

    Two- photon polymerization lithography for optics and photonics: fundamentals, materials, technologies, and applications.Ad- vanced Functional Materials, 33(39):2214211, 2023

    Hao Wang, Wang Zhang, Dimitra Ladika, Haoyi Yu, Dar- ius Gailevi ˇcius, Hongtao Wang, Cheng-Feng Pan, Parvathi Nair Suseela Nair, Yujie Ke, Tomohiro Mori, et al. Two- photon polymerization lithography for optics and photonics: fundamentals, materials, technologies, and applica...

    2023

  62. [70]

    VGGSfM: Visual geometry grounded deep structure from motion

    Jianyuan Wang, Nikita Karaev, Christian Rupprecht, and David Novotny. VGGSfM: Visual geometry grounded deep structure from motion. InProceedings of the IEEE/CVF con- ference on computer vision and pattern recognition, pages 21686–21697, 2024. 2

    2024

  63. [71]

    VGGT: Visual geometry grounded transformer

    Jianyuan Wang, Minghao Chen, Nikita Karaev, Andrea Vedaldi, Christian Rupprecht, and David Novotny. VGGT: Visual geometry grounded transformer. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 5294–5306, 2025. 2, 7

    2025

  64. [72]

    NeuS: Learning neural im- plicit surfaces by volume rendering for multi-view recon- struction.Advances in Neural Information Processing Sys- tems, 34:27171–27183, 2021

    Peng Wang, Lingjie Liu, Yuan Liu, Christian Theobalt, Taku Komura, and Wenping Wang. NeuS: Learning neural im- plicit surfaces by volume rendering for multi-view recon- struction.Advances in Neural Information Processing Sys- tems, 34:27171–27183, 2021. 2, 3, 4, 7, 13

    2021

  65. [73]

    DUSt3R: Geometric 3D vision made easy

    Shuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii, and Jerome Revaud. DUSt3R: Geometric 3D vision made easy. InProceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 20697–20709, 2024. 2

    2024

  66. [74]

    Liv- ing materials fabricated via gradient mineralization of light- inducible biofilms.Nature chemical biology, 17(3):351–359,

    Yanyi Wang, Bolin An, Bin Xue, Jiahua Pu, Xiuli Zhang, Yuanyuan Huang, Yi Yu, Yi Cao, and Chao Zhong. Liv- ing materials fabricated via gradient mineralization of light- inducible biofilms.Nature chemical biology, 17(3):351–359,

  67. [75]

    Synthesis, properties, and multifarious ap- plications of SiC nanoparticles: A review.Ceramics Inter- national, 48(7):8882–8913, 2022

    Yiyuan Wang, Shun Dong, Xiutao Li, Changqing Hong, and Xinghong Zhang. Synthesis, properties, and multifarious ap- plications of SiC nanoparticles: A review.Ceramics Inter- national, 48(7):8882–8913, 2022. 6

    2022

  68. [76]

    NeuS2: Fast learning of neural implicit surfaces for multi-view recon- struction

    Yiming Wang, Qin Han, Marc Habermann, Kostas Dani- ilidis, Christian Theobalt, and Lingjie Liu. NeuS2: Fast learning of neural implicit surfaces for multi-view recon- struction. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 3295–3306, 2023. 3

    2023

  69. [77]

    DeepSFM: Structure from motion via deep bundle adjustment

    Xingkui Wei, Yinda Zhang, Zhuwen Li, Yanwei Fu, and Xi- angyang Xue. DeepSFM: Structure from motion via deep bundle adjustment. InEuropean conference on computer vi- sion, pages 230–247. Springer, 2020. 2

    2020

  70. [78]

    Photometric method for determining 11 surface orientation from multiple images.Optical engineer- ing, 19(1):139–144, 1980

    Robert J Woodham. Photometric method for determining 11 surface orientation from multiple images.Optical engineer- ing, 19(1):139–144, 1980. 2

    1980

  71. [79]

    Recent advances in 3D Gaussian splatting.Computational Visual Media, 10(4): 613–642, 2024

    Tong Wu, Yu-Jie Yuan, Ling-Xiao Zhang, Jie Yang, Yan- Pei Cao, Ling-Qi Yan, and Lin Gao. Recent advances in 3D Gaussian splatting.Computational Visual Media, 10(4): 613–642, 2024. 3

    2024

  72. [80]

    Neural fields in visual computing and beyond.Computer Graphics Forum, 41(2):641–676, 2022

    Yiheng Xie, Towaki Takikawa, Shunsuke Saito, Or Litany, Shiqin Yan, Numair Khan, Federico Tombari, James Tomp- kin, Vincent Sitzmann, and Srinath Sridhar. Neural fields in visual computing and beyond.Computer Graphics Forum, 41(2):641–676, 2022. 3

    2022

  73. [81]

    En- hanced FIB-SEM systems for large-volume 3D imaging

    C Shan Xu, Kenneth J Hayworth, Zhiyuan Lu, Patricia Grob, Ahmed M Hassan, Jos´e G Garc´ıa-Cerd´an, Krishna K Niyogi, Eva Nogales, Richard J Weinberg, and Harald F Hess. En- hanced FIB-SEM systems for large-volume 3D imaging. elife, 6:e25916, 2017. 14

    2017

  74. [82]

    Shang Yan, Aderonke Adegbule, and Tohren C. G. Kibbey. A hybrid 3D SEM reconstruction method optimized for com- plex geologic material surfaces.Micron, 99:26–31, 2017. 2, 4

    2017

  75. [83]

    Fast3R: Towards 3D reconstruction of 1000+ images in one forward pass

    Jianing Yang, Alexander Sax, Kevin J Liang, Mikael Henaff, Hao Tang, Ang Cao, Joyce Chai, Franziska Meier, and Matt Feiszli. Fast3R: Towards 3D reconstruction of 1000+ images in one forward pass. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 21...

  76. [84]

    MVSNet: Depth inference for unstructured multi- view stereo

    Yao Yao, Zixin Luo, Shiwei Li, Tian Fang, and Long Quan. MVSNet: Depth inference for unstructured multi- view stereo. InProceedings of the European conference on computer vision (ECCV), pages 767–783, 2018. 2

    2018

  77. [85]

    V ol- ume rendering of neural implicit surfaces.Advances in Neu- ral Information Processing Systems, 34:4805–4815, 2021

    Lior Yariv, Jiatao Gu, Yoni Kasten, and Yaron Lipman. V ol- ume rendering of neural implicit surfaces.Advances in Neu- ral Information Processing Systems, 34:4805–4815, 2021. 2, 3

    2021

  78. [86]

    Gaussian opacity fields: Efficient adaptive surface reconstruction in unbounded scenes.ACM Transactions on Graphics (ToG), 43(6):1–13, 2024

    Zehao Yu, Torsten Sattler, and Andreas Geiger. Gaussian opacity fields: Efficient adaptive surface reconstruction in unbounded scenes.ACM Transactions on Graphics (ToG), 43(6):1–13, 2024. 3

    2024

  79. [87]

    Managed bumblebees outperform honeybees in increasing peach fruit set in China: different limiting pro- cesses with different pollinators.Plos one, 10(3):e0121143,

    Hong Zhang, Jiaxing Huang, Paul H Williams, Bernard E Vaissi`ere, Zhiyong Zhou, Qinbao Gai, Jie Dong, and Jian- dong An. Managed bumblebees outperform honeybees in increasing peach fruit set in China: different limiting pro- cesses with different pollinators.Plos one, 10(3):e0121143,

  80. [88]

    Superalloys fracture process inference based on overlap analysis of 3D models.Communications Engineering, 3(1):108, 2024

    Xuecheng Zhang, Guanghao Guo, Zixin Li, Wenchao Meng, Yuefei Zhang, Qing Ye, Jin Wang, Shibo He, Xinbao Zhao, Jiming Chen, et al. Superalloys fracture process inference based on overlap analysis of 3D models.Communications Engineering, 3(1):108, 2024. 2, 3, 4, 6

    2024

  81. [89]

    GeoMVSNet: Learning multi-view stereo with geometry perception

    Zhe Zhang, Rui Peng, Yuxi Hu, and Ronggang Wang. GeoMVSNet: Learning multi-view stereo with geometry perception. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 21508– 21518, 2023. 2, 15, 16, 17

    2023

  82. [90]

    Additive manufacturing of silica aerogels.Nature, 584 (7821):387–392, 2020

    Shanyu Zhao, Gilberto Siqueira, Sarka Drdova, David Nor- ris, Christopher Ubert, Anne Bonnin, Sandra Galmarini, Michal Ganobjak, Zhengyuan Pan, Samuel Brunner, et al. Additive manufacturing of silica aerogels.Nature, 584 (7821):387–392, 2020. 1 12 C. Sample Preparation C.1. TP...

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.