Pith. sign in

REVIEW 2 major objections 1 minor 43 references

Towards terahertz nanomechanics

T0 review · 2 major / 1 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Suspended Lamb-wave resonators in 67-nanometer lithium niobate reach 220 GHz, doubling the prior record.

desk verdict A record-frequency claim in 67 nm lithium niobate that I can't verify from the supplied text—worth a referee's time, but the abstract alone can't carry it. read the letter →

arxiv 2508.03933 v1 pith:D3ZSB37N submitted 2025-08-05 physics.app-ph

Jiacheng Xie , Weifeng Wu , Mohan Shen , Patrick Fay , Hong X. Tang This is my paper · ORCID
classification physics.app-ph
keywords terahertznanomechanicsLamb-waveresonatorslithiumniobatethin-filmquantumphononicsacousticlossesresonantfrequency
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 pushes electromechanical resonators to 220 GHz by thinning lithium niobate to 67 nanometers, doubling the previous record and approaching the terahertz threshold. The frequency gain comes from the thickness-wavelength scaling of Lamb waves in a suspended membrane. The authors show the tradeoff: sub-100-nanometer films suffer higher acoustic losses, so reaching terahertz will require controlling surface defects.

What carries the argument

The central object is a suspended Lamb-wave resonator in a piezoelectric thin film; its resonant frequency is set by the film thickness, which fixes the acoustic wavelength. Thinning lithium niobate from 300 nm to 67 nm shrinks the wavelength to tens of nanometers and lifts the resonance from tens of gigahertz to 220 GHz, with the suspended membrane keeping the mode confined and electrically addressable.

What would settle it

Look for the resonance peak in the transmission spectrum of the 67-nanometer device and check whether its frequency moves with film thickness according to the Lamb-wave dispersion relation across all three measured thicknesses, and whether it disappears when the film is not freely suspended.

Watch

Extended reading notes

Core claim

The paper reports suspended Lamb-wave resonators fabricated in 67-nanometer-thick lithium niobate that reach resonant frequencies up to 220 GHz—twice the previous record and within a factor of five of the terahertz band. By stepping the film thickness down from 300 nm through several stages, the authors demonstrate that aggressive thinning raises the resonance frequency as expected from the acoustic-wavelength scaling, while also showing that acoustic losses increase in sub-100-nanometer films. The authors argue that this combination sets the path for terahertz nanomechanics: further frequency gains will depend on mitigating surface defects in the thinnest films.

Load-bearing premise

The central claim depends on the electrical signatures at 220 GHz being genuine mechanical Lamb-wave resonances of the 67-nanometer suspended film, rather than electromagnetic feedthrough, calibration artifacts, or spurious electrical resonances.

Editorial extensions

If this is right

  • A 220 GHz mechanical resonance places electromechanical devices in the sub-terahertz band, opening wide bandwidths for phononic signal processing.
  • At these frequencies, a mechanical resonator can remain in its quantum ground state at liquid-helium temperatures, avoiding the millikelvin cooling that gigahertz resonators require.
  • The demonstrated thickness-frequency scaling implies that thinning toward 30-nanometer films could reach the terahertz range, if losses can be controlled.
  • Observed acoustic losses in sub-100-nanometer films point to surface defects as the main obstacle, making surface-quality engineering the critical next step.

Reading between the lines

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

  • If the same scaling holds to a few tens of nanometers, the lithium niobate platform could host several-hundred-gigahertz resonators practical for coupling to optical photons or superconducting qubits.
  • The loss trend reported here suggests that surface smoothing or passivation layers could be a direct route to recovering quality factor in ultrathin films.
  • The work makes piezoelectric Lamb-wave resonators a leading candidate for electrically actuated terahertz mechanical oscillators, complementing optically driven approaches.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 1 minor

Summary. The paper, as presented, claims that suspended Lamb-wave resonators fabricated in 67-nm-thick lithium niobate achieve resonant frequencies up to 220 GHz, reportedly doubling the previous record and approaching the terahertz range. The abstract states that aggressive film thinning from 300 nm to 67 nm increases frequency and that sub-100 nm films exhibit increased acoustic losses. These claims are presented entirely at the abstract level: the supplied full text is an unrelated manuscript on AI red-teaming ('ASTRA') and contains no fabrication details, measurement data, de-embedding procedures, Q-factor extraction, or comparison analysis for the nanomechanical results. Thus, the central result cannot be inspected or verified from the submitted material.

Significance. If the stated result were fully substantiated, it would be highly significant: a 220 GHz electromechanical resonator in a 67 nm piezoelectric film would be the highest-frequency mechanical resonator reported to date, with implications for phononic signal processing and for operating in the quantum ground state at readily accessible cryogenic temperatures. The abstract's claims are falsifiable and concrete, which is a strength, but the manuscript currently provides no evidence, reproducible code, or derivations to support them. The supplied full text is not the paper described by the abstract, so the significance of the reported advance cannot be assessed from the manuscript in its present form.

major comments (2)
  1. [Full text (supplied manuscript body)] The body of arXiv:2508.03933 is not the paper described by the abstract. It is an unrelated computer-security manuscript titled 'ASTRA: Autonomous Spatial-Temporal Red-teaming for AI Software Assistants.' Consequently, none of the central claims—67-nm lithium niobate membranes, resonant frequencies up to 220 GHz, doubling of the prior record, or increased acoustic losses in sub-100 nm films—has any supporting derivation, measurement, or analysis in the submitted text. This is a load-bearing omission: the central claim is currently an abstract-level assertion only, and no part of the claimed result can be inspected.
  2. [Abstract (measurement substantiation)] The abstract reports 'resonant frequencies as high as 220 GHz' but provides no measurement details. In particular, there is no description of how the electrical response was de-embedded, no S-parameter or transmission data, no method for extracting the resonance frequency or Q-factor, and no mode-shape or dispersion identification to distinguish a Lamb-wave resonance of the 67 nm film from electromagnetic feedthrough or a spurious electrical resonance. Because the acoustic wavelength at 220 GHz would be on the order of tens of nanometers, this identification is nontrivial and is required to substantiate the claim. Also absent is a citation or comparative data identifying the previous record that is allegedly doubled.
minor comments (1)
  1. [Abstract (presentation)] The abstract states the film is thinned 'through several stages' without specifying the intermediate thicknesses; it would improve clarity to state the exact thickness series. In addition, specify whether 220 GHz is the highest measured resonance peak or the center frequency of a fitted response, and provide a reference to the prior record.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the abstract reports measured frequencies; no fitted parameter, self-citation chain, or constructed identity is visible.

full rationale

The claimed derivation chain in the abstract is minimal: the authors thinned lithium niobate from 300 nm to 67 nm, fabricated suspended Lamb-wave resonators, and report resonant frequencies up to 220 GHz. The headline number is presented as a measurement outcome, not as the output of a fitted model, a uniqueness argument, or a definitional identity. No equation in the visible text ties 220 GHz to an input parameter, and no prior result by the same authors is invoked to force the conclusion. The comparison to a previous record is an external benchmark claim, not a circular reduction. The only substantive concern raised by the reader—that the 220 GHz electrical signatures may be feedthrough or spurious resonances rather than genuine Lamb modes—is a measurement-validity or evidence question, not a circularity question under the specified patterns. Moreover, the supplied full text is an unrelated paper (ASTRA, arXiv:2508.03936, on LLM red-teaming), so no methods, S-parameters, de-embedding steps, or mode-identification details from the terahertz paper can be inspected. That prevents any exhibit of a specific reduction, which hard rule 1 requires. Accordingly, the honest finding is no significant circularity: score 0.

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

This ledger is necessarily shallow because only the abstract is available; the supplied body text is a different paper. No free parameters or invented entities appear at the abstract level. The two axioms listed are the interpretive premises that any frequency record claim depends on: that the observed signature is a real acoustic mode, and that the expected scaling physics applies.

assumptions (2)
  • domain assumption The measured electrical responses at 220 GHz correspond to genuine acoustic Lamb-wave resonances of the suspended film rather than electromagnetic feedthrough or other artifacts.
    The central claim depends on correct mode identification and de-embedding at frequencies where parasitic electrical coupling is severe; this cannot be checked from the abstract.
  • domain assumption Frequency scaling with film thickness follows the expected acoustic dispersion, with the thinned film acting as the acoustic-wavelength-defining dimension.
    The strategy of thinning film to the acoustic wavelength presupposes the standard thickness-mode scaling of Lamb waves; the abstract does not present the dispersion data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards terahertz nanomechanics." pith.science (2026). https://pith.science/paper/D3ZSB37N

@misc{pith2026250803933,
  author       = {Pith},
  title        = {Pith review of: Towards terahertz nanomechanics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D3ZSB37N}},
  note         = {Machine review of arXiv:2508.03933}
}
read the original abstract

Advancing electromechanical resonators towards terahertz frequencies opens vast bandwidths for phononic signal processing. In quantum phononics, mechanical resonators at these frequencies can remain in their quantum ground state even at kelvin temperatures, obviating the need for millikelvin cooling typically required for GHz resonators. However, electrical actuation and detection of mechanical motion at such high frequencies present significant challenges, primarily due to the need for device miniaturization to support acoustic waves with nanometer-scale wavelengths. One effective strategy is to aggressively thin down piezoelectric thin films, ideally to a thickness on the order of the acoustic wavelength, which is in the tens of nanometers. In this work, we aggressively reduce the thickness of lithium niobate from 300 nm to 67 nm through several stages, and fabricate suspended Lamb-wave resonators at each thickness level. These resonators achieve resonant frequencies as high as 220 GHz, doubling the previous record and approaching the terahertz frequency threshold. While ultrathin films exhibit a clear advantage in frequency gains, they also experience increased acoustic losses. Our results suggest that future advances in terahertz nanomechanics will critically rely on mitigating surface defects in sub-100 nm thin films.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 21 canonical work pages

  1. [1]

    https://repoaudit-home.github.io/index.html,

    Repoaudit: Auditing code as human. https://repoaudit-home.github.io/index.html,

    work page

  2. [2]

    Code Review Tool: Amazon CodeGuru Security

    Amazon. Code Review Tool: Amazon CodeGuru Security. https://aws.amazon.com/ codeguru/, 2025. [Online; accessed 4-May-2025]

    work page 2025

  3. [3]

    Purple llama cyberseceval: A secure coding benchmark for language models

    Manish Bhatt, Sahana Chennabasappa, Cyrus Nikolaidis, Shengye Wan, Ivan Evtimov, Do- minik Gabi, Daniel Song, Faizan Ahmad, Cornelius Aschermann, Lorenzo Fontana, et al. Purple llama cyberseceval: A secure coding benchmark for language models. arXiv preprint arXiv:2312.04724, 2023

    arXiv 2023

  4. [4]

    Jailbreaking black box large language models in twenty queries

    Patrick Chao, Alexander Robey, Edgar Dobriban, Hamed Hassani, George J Pappas, and Eric Wong. Jailbreaking black box large language models in twenty queries. arXiv preprint arXiv:2310.08419, 2023

    arXiv 2023

  5. [5]

    When LLM Meets DRL: Advancing Jailbreaking Efficiency via DRL-guided Search

    Xuan Chen, Yuzhou Nie, Wenbo Guo, and Xiangyu Zhang. When llm meets drl: Advancing jailbreaking efficiency via drl-guided search. arXiv preprint arXiv:2406.08705, 2024

    work page Pith review arXiv 2024

  6. [6]

    Masterkey: Automated jailbreak across multiple large language model chatbots

    Gelei Deng, Yi Liu, Yuekang Li, Kailong Wang, Ying Zhang, Zefeng Li, Haoyu Wang, Tianwei Zhang, and Yang Liu. Masterkey: Automated jailbreak across multiple large language model chatbots. arXiv preprint arXiv:2307.08715, 2023

    arXiv 2023

  7. [7]

    A wolf in sheep’s clothing: Generalized nested jailbreak prompts can fool large language models easily

    Peng Ding, Jun Kuang, Dan Ma, Xuezhi Cao, Yunsen Xian, Jiajun Chen, and Shujian Huang. A wolf in sheep’s clothing: Generalized nested jailbreak prompts can fool large language models easily. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (V olume1: Long Pape...

    work page 2024

  8. [8]

    The llama 3 herd of models

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783, 2024

    arXiv 2024

Show all 43 references
  1. [9]

    How ai is being audited—and why it matters, 2024

    Financial Times. How ai is being audited—and why it matters, 2024. https://www.ft.com/ content/8a54932d-d9a9-4a69-969d-89d8b2de149f

    2024

  2. [10]

    Stochastic relaxation, gibbs distributions, and the bayesian restoration of images

    Stuart Geman and Donald Geman. Stochastic relaxation, gibbs distributions, and the bayesian restoration of images. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI- 6(6):721–741, 1984

    1984

  3. [11]

    Deliberative alignment: Reasoning enables safer language models

    Melody Y Guan, Manas Joglekar, Eric Wallace, Saachi Jain, Boaz Barak, Alec Helyar, Rachel Dias, Andrea Vallone, Hongyu Ren, Jason Wei, et al. Deliberative alignment: Reasoning enables safer language models. arXiv preprint arXiv:2412.16339, 2024

    2024 arXiv

  4. [12]

    compe- tency

    Divij Handa, Zehua Zhang, Amir Saeidi, Shrinidhi Kumbhar, and Chitta Baral. When" compe- tency" in reasoning opens the door to vulnerability: Jailbreaking llms via novel complex ciphers. arXiv preprint arXiv:2402.10601, 2024

    2024

  5. [13]

    Large language models for code: Security hardening and adversarial testing

    Jingxuan He and Martin Vechev. Large language models for code: Security hardening and adversarial testing. In Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security, pages 1865–1879, 2023

    2023

  6. [14]

    Artprompt: Ascii art-based jailbreak attacks against aligned llms

    Fengqing Jiang, Zhangchen Xu, Luyao Niu, Zhen Xiang, Bhaskar Ramasubramanian, Bo Li, and Radha Poovendran. Artprompt: Ascii art-based jailbreak attacks against aligned llms. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (V olume1: L...

    2024

  7. [15]

    Red queen: Safeguarding large language models against concealed multi-turn jailbreaking

    Yifan Jiang, Kriti Aggarwal, Tanmay Laud, Kashif Munir, Jay Pujara, and Subhabrata Mukher- jee. Red queen: Safeguarding large language models against concealed multi-turn jailbreaking. arXiv preprint arXiv:2409.17458, 2024

    2024 arXiv

  8. [16]

    Multiverse: Exposing large language model alignment problems in diverse worlds

    Xiaolong Jin, Zhuo Zhang, and Xiangyu Zhang. Multiverse: Exposing large language model alignment problems in diverse worlds. arXiv preprint arXiv:2402.01706, 2024. 21

    2024 arXiv

  9. [17]

    Drattack: Prompt decomposition and reconstruction makes powerful llm jailbreakers

    Xirui Li, Ruochen Wang, Minhao Cheng, Tianyi Zhou, and Cho-Jui Hsieh. Drattack: Prompt decomposition and reconstruction makes powerful llm jailbreakers. arXiv preprint arXiv:2402.16914, 2024

    2024 arXiv

  10. [18]

    The reversal curse: Llms trained on ‘a is b’ fail to learn ‘b is a’.arXiv preprint arXiv:2305.13283, 2023

    Xisen Li, Jiefu Liu, Chunting Zhang, Colin Raffel, Kristina Tau, James Zou, and Dan Jurafsky. The reversal curse: Llms trained on ‘a is b’ fail to learn ‘b is a’.arXiv preprint arXiv:2305.13283, 2023

    2023 arXiv

  11. [19]

    Deepinception: Hypnotize large language model to be jailbreaker

    Xuan Li, Zhanke Zhou, Jianing Zhu, Jiangchao Yao, Tongliang Liu, and Bo Han. Deepinception: Hypnotize large language model to be jailbreaker. arXiv preprint arXiv:2311.03191, 2023

    2023 arXiv

  12. [20]

    Making them ask and answer: Jailbreaking large language models in few queries via disguise and reconstruction

    Tong Liu, Yingjie Zhang, Zhe Zhao, Yinpeng Dong, Guozhu Meng, and Kai Chen. Making them ask and answer: Jailbreaking large language models in few queries via disguise and reconstruction. In 33rd USENIX Security Symposium (USENIX Security 24), pages 4711– 4728, 2024

    2024

  13. [21]

    Autodan: Generating stealthy jailbreak prompts on aligned large language models

    Xiaogeng Liu, Nan Xu, Muhao Chen, and Chaowei Xiao. Autodan: Generating stealthy jailbreak prompts on aligned large language models. In The Twelfth International Conference on Learning Representations, 2024

    2024

  14. [22]

    Flipattack: Jailbreak llms via flipping

    Yue Liu, Xiaoxin He, Miao Xiong, Jinlan Fu, Shumin Deng, and Bryan Hooi. Flipattack: Jailbreak llms via flipping. arXiv preprint arXiv:2410.02832, 2024

    2024 arXiv

  15. [23]

    Codechameleon: Personalized encryption framework for jailbreaking large language models

    Huijie Lv, Xiao Wang, Yuansen Zhang, Caishuang Huang, Shihan Dou, Junjie Ye, Tao Gui, Qi Zhang, and Xuanjing Huang. Codechameleon: Personalized encryption framework for jailbreaking large language models. arXiv preprint arXiv:2402.16717, 2024

    2024 arXiv

  16. [24]

    Ai in software market size, share & trends analysis report, 2023–2033, 2024

    Market.US. Ai in software market size, share & trends analysis report, 2023–2033, 2024. https://market.us/report/ai-in-software-market/

    2023

  17. [25]

    Tree of attacks: Jailbreaking black-box llms automatically.Advances in Neural Information Processing Systems, 37:61065–61105, 2024

    Anay Mehrotra, Manolis Zampetakis, Paul Kassianik, Blaine Nelson, Hyrum Anderson, Yaron Singer, and Amin Karbasi. Tree of attacks: Jailbreaking black-box llms automatically.Advances in Neural Information Processing Systems, 37:61065–61105, 2024

    2024

  18. [26]

    Mitre att&ck framework

    MITRE Corporation. Mitre att&ck framework. https://attack.mitre.org/, 2024. Ac- cessed: 2025-05-18

    2024

  19. [27]

    Allen Newell and Herbert A. Simon. Human Problem Solving. Prentice-Hall, Englewood Cliffs, NJ, 1972

    1972

  20. [28]

    Safety alignment should be made more than just a few tokens deep

    Xiangyu Qi, Ashwinee Panda, Kaifeng Lyu, Xiao Ma, Subhrajit Roy, Ahmad Beirami, Prateek Mittal, and Peter Henderson. Safety alignment should be made more than just a few tokens deep. In ICLR, 2025

    2025

  21. [29]

    Derail yourself: Multi-turn llm jailbreak attack through self- discovered clues

    Qibing Ren, Hao Li, Dongrui Liu, Zhanxu Xie, Xiaoya Lu, Yu Qiao, Lei Sha, Junchi Yan, Lizhuang Ma, and Jing Shao. Derail yourself: Multi-turn llm jailbreak attack through self- discovered clues. arXiv preprint arXiv:2410.10700, 2024

    2024

  22. [30]

    Amazon nova ai challenge, trusted ai: Advancing secure, ai-assisted software development

    Sattvik Sahai, Prasoon Goyal, Michael Johnston, Anna Gottardi, Yao Lu, Lucy Hu, Luke Dai, Shaohua Liu, Samyuth Sagi, Hangjie Shi, Desheng Zhang, Lavina Vaz, Leslie Ball, Maureen Murray, Rahul Gupta, and Shankar Ananthakrishnan. Amazon nova ai challenge, trusted ai: Advancing s...

    2025

  23. [31]

    Pal: Proxy-guided black-box attack on large language models

    Chawin Sitawarin, Norman Mu, David Wagner, and Alexandre Araujo. Pal: Proxy-guided black-box attack on large language models. arXiv preprint arXiv:2402.09674, 2024

    2024 arXiv

  24. [32]

    All in how you ask for it: Simple black-box method for jailbreak attacks

    Kazuhiro Takemoto. All in how you ask for it: Simple black-box method for jailbreak attacks. Applied Sciences, 14(9):3558, 2024

    2024

  25. [33]

    Cognitive overload: Jailbreaking large language models with overloaded logical thinking

    Nan Xu, Fei Wang, Ben Zhou, Bang Zheng Li, Chaowei Xiao, and Muhao Chen. Cognitive overload: Jailbreaking large language models with overloaded logical thinking. arXiv preprint arXiv:2311.09827, 2023. 22

    2023 arXiv

  26. [34]

    Prosec: Fortifying code llms with proactive security alignment

    Xiangzhe Xu, Zian Su, Jinyao Guo, Kaiyuan Zhang, Zhenting Wang, and Xiangyu Zhang. Prosec: Fortifying code llms with proactive security alignment. arXiv preprint arXiv:2411.12882, 2024

    2024 arXiv

  27. [35]

    Jigsaw puzzles: Splitting harmful questions to jailbreak large language models

    Hao Yang, Lizhen Qu, Ehsan Shareghi, and Gholamreza Haffari. Jigsaw puzzles: Splitting harmful questions to jailbreak large language models. arXiv preprint arXiv:2410.11459, 2024

    2024 arXiv

  28. [36]

    Chain of attack: a semantic-driven contextual multi-turn attacker for llm

    Xikang Yang, Xuehai Tang, Songlin Hu, and Jizhong Han. Chain of attack: a semantic-driven contextual multi-turn attacker for llm. arXiv preprint arXiv:2405.05610, 2024

    2024 arXiv

  29. [37]

    Fuzzllm: A novel and universal fuzzing framework for proactively discovering jailbreak vulnerabilities in large language mod- els

    Dongyu Yao, Jianshu Zhang, Ian G Harris, and Marcel Carlsson. Fuzzllm: A novel and universal fuzzing framework for proactively discovering jailbreak vulnerabilities in large language mod- els. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Pr...

    2024

  30. [38]

    Gptfuzzer: Red teaming large language models with auto-generated jailbreak prompts

    Jiahao Yu, Xingwei Lin, Zheng Yu, and Xinyu Xing. Gptfuzzer: Red teaming large language models with auto-generated jailbreak prompts. arXiv preprint arXiv:2309.10253, 2023

    2023 arXiv

  31. [39]

    How johnny can persuade llms to jailbreak them: Rethinking persuasion to challenge ai safety by human- izing llms

    Yi Zeng, Hongpeng Lin, Jingwen Zhang, Diyi Yang, Ruoxi Jia, and Weiyan Shi. How johnny can persuade llms to jailbreak them: Rethinking persuasion to challenge ai safety by human- izing llms. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguis...

    2024

  32. [40]

    Make them spill the beans! coercive knowledge extraction from (production) llms, 2023

    Zhuo Zhang, Guangyu Shen, Guanhong Tao, Siyuan Cheng, and Xiangyu Zhang. Make them spill the beans! coercive knowledge extraction from (production) llms, 2023

    2023

  33. [41]

    Improving llm safety alignment with dual-objective optimization

    Xuandong Zhao, Will Cai, Tianneng Shi, David Huang, Licong Lin, Song Mei, and Dawn Song. Improving llm safety alignment with dual-objective optimization. arXiv preprint arXiv:2503.03710, 2025

    2025

  34. [42]

    Improving alignment and robustness with circuit breakers

    Andy Zou, Long Phan, Justin Wang, Derek Duenas, Maxwell Lin, Maksym Andriushchenko, J Zico Kolter, Matt Fredrikson, and Dan Hendrycks. Improving alignment and robustness with circuit breakers. In The Thirty-eighth Annual Conference on Neural Information Processing Systems, 202...

    2024

  35. [2025]

    Accessed: 2025-04-24

    2025

Pith tools

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