Pith. sign in

REVIEW 3 major objections 5 minor 35 references

Post-Hoc Trajectory-Risk Certification for Modular LLM-Based Security Agents

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read For frozen multi-stage LLM security pipelines, trajectory-level miscoverage can be certified post-hoc from per-stage audits plus pairwise failure-overlap bounds.

desk verdict Solid post-hoc certification theory and honest self-correction, but the headline 13.7% empirical gain is built on an audit size chosen after a pilot look and is not a valid fixed-n certificate as reported. read the letter →

arxiv 2608.05199 v1 pith:47WICBYC submitted 2026-08-04 cs.CR cs.AI

Zhenpeng Li This is my paper · ORCID
classification cs.CRcs.AI
keywords trajectorycoverageconformalpredictionmulti-stageLLMagentsintrusiondetectionpost-hoccertificationsamplecomplexityspanning-treeboundlabelnestingartifact
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

The paper establishes a way to turn per-step safety certificates of a frozen, multi-stage LLM security pipeline into a guarantee on the whole trajectory, using only audit data and no joint recalibration. The central claim is that trajectory miscoverage is bounded above by the sum of certified per-step risk bounds minus a certified pairwise failure-overlap gain, provided the gain is taken over a spanning tree rather than by naive inclusion–exclusion; the naive extension to three or more stages is a lower bound, not an upper bound. A second central claim is that certifying any positive dependence gain requires an audit sample of order $1/q_{12}$, where $q_{12}$ is the joint-failure rate, and the certificate can flip from worse than the Bonferroni bound to tighter than it purely as the audit reaches that scale. The paper also shows that a coarse-to-fine label design can mechanically manufacture near-perfect measured inter-stage correlation, and that correcting this artifact reveals task-dependent coupling attributable mostly to shared sample difficulty.

What carries the argument

The spanning-tree pairwise upper bound is the load-bearing object: for any spanning tree $T$ on the $K$ stages, trajectory miscoverage is at most $\sum_k p_k - \sum_{(i,j)\in T} q_{ij}$, where $p_k$ is the per-step marginal risk and $q_{ij}$ is the pairwise failure-overlap probability. This bound is what makes the certificate valid for every $K$, replacing the natural pairwise inclusion–exclusion truncation, which is a lower bound once $K>2$. The certificate assembles the bound from simultaneous concentration or exact binomial confidence bounds on the $p_k$ and $q_{ij}$, and the maximizing tree may be selected after seeing the audit data. A companion diagnostic result shows that if a fine label is a deterministic function of a coarse label, the two stages’ failure events are nested and their measured correlation is forced near one by construction.

What would settle it

Simulate a realized two-stage chain with known per-step risks $p_1,p_2$ and overlap $q_{12}$, draw audit samples at the size prescribed by the paper’s threshold, and check whether the certificate $\min\{1, U_1+U_2-L_{12}\}$ holds with empirical frequency at least $1-\delta_p-\delta_q$; if the frequency falls below the claimed confidence, the post-hoc certificate is invalid.

Watch

Extended reading notes

Core claim

Condition on the realized trained-and-calibrated chain $\theta$. If an audit of $n$ i.i.d. trajectories yields simultaneous upper bounds $U_k$ on each stage’s miscoverage probability and lower bounds $L_{ij}$ on each pair’s failure overlap, then with probability at least $1-\delta_p-\delta_q$ the true trajectory miscoverage satisfies $R(\theta) \le \min\{1, \sum_k U_k - \max_T \sum_{(i,j)\in T} L_{ij}\}$, where $T$ runs over spanning trees on the stages. The data-dependent choice of the best tree after seeing the audit does not break validity, because the underlying spanning-tree bound holds for every tree simultaneously. On a two-stage intrusion-detection pipeline over six open LLMs and two datasets, a direct one-sided binomial audit of the union event is 13.7% tighter than the Bonferroni bound once the audit reaches the sample size the theory requires, while the modular certificate that uses only per-stage certificates and one pairwise overlap recovers a 0.6% average gain; at an under-sized audit the same certificate is worse than Bonferroni, matching the derived threshold.

Load-bearing premise

The load-bearing premise is that the audit sample is drawn independently from the exact deployment distribution of the frozen chain, so the audit’s empirical failure rates stay representative of real traffic.

Editorial extensions

If this is right

  • Operators of frozen pipelines can certify trajectory-level risk from audit data alone, without retraining or joint recalibration, whenever per-stage certificates and pairwise overlap lower bounds are obtainable.
  • An audit too small for the joint-failure rate makes dependence-aware certificates worse than the Bonferroni bound; the paper makes the required sample size explicit as $\Theta(1/q_{12})$ for detecting any positive gain and $\Theta(1/\gamma^2)$ for beating the nominal target.
  • For pipelines with three or more stages, pairwise information alone cannot determine trajectory risk; the spanning-tree bound remains a valid upper bound, but tighter two-sided statements require higher-order overlap data.
  • Positive inter-stage error correlation is an asset for the union-type trajectory failure criterion; the same dependence would be a liability for an intersection-type redundant-system criterion.
  • Evaluations of multi-stage LLM agents must check that stage labels are not deterministic refinements of earlier stage labels, or measured inter-stage dependence can be a labeling artifact rather than a property of the model.

Reading between the lines

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

  • The paper’s own comparison implies that when full joint audit vectors are available at $K=2$, operators should deploy the direct one-sided binomial bound on the union event rather than the modular decomposition; the modular certificate is the useful object only when only pairwise summaries exist or when $K>2$.
  • The $\Theta(1/q_{12})$ audit threshold suggests that for very rare joint failures, dependence-aware certification is impractical; an adaptive or time-uniform sequential audit that stops as soon as the overlap lower bound turns positive would make the procedure deployable.
  • The label-nesting theorem likely generalizes to any hierarchical taxonomy: whenever later stages refine earlier labels deterministically or near-deterministically, reported inter-stage correlation is inflated; this gives a testable checklist for other multi-stage evaluation pipelines.
  • The near-one empty-set rate under cross-dataset shift suggests an online, label-free drift alarm: monitoring the rate at which conformal prediction sets are empty or singleton could detect exchangeability failure before any ground truth arrives.
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

3 major / 5 minor

Summary. The paper formalizes trajectory-level risk certification for modular LLM-based security pipelines whose stages are individually calibrated but cannot be jointly recalibrated. The main theory is centered on a post-hoc certificate (Theorem 7) that combines simultaneous upper bounds on per-step miscoverage probabilities with lower bounds on pairwise failure overlaps via a spanning-tree bound, valid even when the tree is selected from the data. The paper shows that the natural K=2 inclusion-exclusion identity does not extend to K>=3 as an upper bound, proves a non-identifiability result for pairwise information once K>=3, derives fixed-n sample-complexity bounds for certifying positive dependence and for beating the nominal Bonferroni target, and proves that certain coarse-to-fine label mappings mechanically induce near-perfect measured correlation. The empirical study on a two-stage intrusion-detection pipeline across six LLMs and two datasets reports a 13.7% tighter direct union audit than Bonferroni at a boosted audit scale, a smaller 0.6% average gain for the modular decomposed certificate, evidence that residual coupling reflects shared sample difficulty rather than shared model representations, and a cross-dataset shift experiment in which single-step coverage collapses to zero despite high raw accuracy. The paper is unusually candid about its own earlier deterministic label-mapping artifact and about the limitations of pairwise information for K>=3.

Significance. If the results hold, the paper makes a useful conceptual contribution to conformal and risk-control composition: it separates marginal from realized-chain risk, distinguishes structural from certifiable dependence gains, supplies a valid K>=2 spanning-tree certificate with data-dependent tree selection, and gives rate-matching upper and lower bounds on the audit size needed to certify positive dependence. The main-text proofs that are included (notably Theorem 3, Theorem 6, Theorem 7, Theorem 8, Theorem 10, Theorem 13, Theorem 14, Theorem 16) are clear and internally sound. The paper also deserves credit for explicitly retracting its earlier near-1 correlation finding as a labeling artifact and for designing a same-model/cross-model/permuted protocol that separates shared-difficulty from representation-sharing accounts. The central theoretical framework is defensible, but the headline empirical certificate is compromised by an adaptive audit-size selection that the paper does not validly justify, and several substantive proofs are deferred to a supplementary file not present in this arXiv version.

major comments (3)
  1. [Section VI-D, paragraphs on the boosted audit scale] The headline numerical claim, that the direct union audit U_union becomes 13.7% tighter than Bonferroni (0.086 vs 0.100), is computed at an audit sample size (3,500/1,376) that was chosen after inspecting a smaller pilot audit (397/109) in which the union bound was worse than Bonferroni and some configurations had zero joint failures. Theorems 7 and 10 are fixed-n statements: they provide valid simultaneous confidence bounds for a pre-specified n, but they do not cover an n selected after seeing pilot data unless the final audit pool is independent of the pilot and the scaling rule was fixed in advance. The text does not state that the boosted pool is disjoint from the pilot pool, does not report a pre-registered or otherwise pre-specified scaling rule, and does not use a time-uniform confidence sequence even though Section II-C explicitly identifies those sequences as the appropriate tool for data-dependent audit sizes. As written, the 13.7% gain and the 7-of-12 count of configurations below 0.10 are not valid 1-delta certificates; they should be re-derived with an always-valid bound, an independent pre-specified split, or clearly re-labeled as an exploratory observation rather than a certified gain.
  2. [Section IV-C and Section IV-F, deferred proofs] Theorem 5, Theorem 9, Theorem 12, Proposition 2, and Theorem 17 are stated in the main text, but their proofs are explicitly deferred to a supplementary material that is not included in this arXiv version. These results carry load-bearing content: the second-order degeneracy threshold, the finite-sample certifiability gap, the sample-complexity lower bound for beating the nominal target, the coupling provenance decomposition, and the dependence duality. The reader cannot currently verify these claims from the manuscript, and the claim of matching upper and information-theoretic lower bounds depends in part on these missing derivations. The revised version should include the full proofs in the paper or in an available supplement, or clearly state where the supplement can be obtained.
  3. [Equation (12) and Remark 1] The plug-in estimate b_BIE in Eq. (12) replaces the true per-step miscoverage probabilities p1 and p2 in Theorem 3 with the nominal budgets alpha1 and alpha2, even though the theorem's identity is written in terms of p1 and p2. The paper justifies this informally by saying the per-step conformal sets are approximately saturated, but the approximation is not quantified and the plug-in is then used in comparative tables. Please make explicit that Eq. (12) is only valid as an approximation when the achieved miscoverage rates are close to their budgets, or replace alpha_k by estimates of p_k and report the effect on the residual analysis.
minor comments (5)
  1. [Introduction] The phrase 'chainstraffic classification' near the start of Section I should be corrected to 'chain traffic classification'.
  2. [Table XI caption] The caption contains '1OF12 CONFIGURATIONS' without a space; it should read '1 of 12 configurations'.
  3. [Section VI-D, first paragraphs] The notation for the pilot audit size (pool n=497/139, test-audit 397/109) is not connected to the Ncal/Ntest values in Table I; please define the calibration sizes used in the pilot and state how the pilot pool relates to the final pool, since this is exactly the adaptive-selection issue in the major comment above.
  4. [Section VI-B] The sentence reporting mean |C1(x)|=0.964 'of 5 categories' should say 'mean cardinality of the prediction set across samples' to avoid the impression that set sizes are fractional; the subsequent explanation about singleton rates clarifies this, but the wording could be tightened.
  5. [Section VI-D, Remark 3 reference] Remark 3 says U_union < U1+U2-L12 holds in every one of the 12 configurations, but the paper does not give a table of the decomposed versus direct values per configuration; please add the per-configuration comparison or state where it can be found.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: derivations are self-contained; the audit-size scaling in Section VI-D is a selection-validity concern, not a circular step.

full rationale

All theorems are proved from external, standard results: Boole's inequality and the chain rule (Theorem 1), inclusion-exclusion and Pearson-correlation algebra (Theorem 3), Bonferroni's parity inequalities (Theorem 4), Hunter's spanning-tree union bound cited as [25] (Theorem 6), Hoeffding and Clopper-Pearson intervals (Theorems 7-10), Bretagnolle-Huber testing (Theorems 11-12), explicit atom constructions (Theorems 13-14), and the law of total covariance (Proposition 2). No term is defined in terms of the quantity it is used to certify; no fitted constant is relabeled as a prediction; and no load-bearing premise rests on a self-citation. The only self-referential element is the audit-size scale-up in Section VI-D: the fixed-n certificate in Equation (39) is valid for a pre-specified n, whereas the boosted audit size n=3,500/1,376 was adopted after a pilot audit showed U_union=0.124 and after estimating q12 from that pilot. This is a statistical selection-validity concern about the empirical demonstration, not an equation-level circularity; the paper itself notes in Section II-C that time-uniform confidence sequences are the appropriate tool when audit size is data-dependent. The reported 13.7% tightness is an observation at the selected scale, not a quantity forced by the theorem's algebra.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central certificates rest on standard inequalities and on the i.i.d. audit assumption; no free constants are fitted to make the theorems true, and no new physical or model entities are introduced. The empirical plug-in correlation is an estimate, not a parameter of the derivation.

assumptions (5)
  • domain assumption Audit sample A_n is i.i.d. from the deployment distribution of the realized chain theta, and the audit is independent of threshold selection (superpopulation model).
    Entered in Section VI-D when test samples are modeled as i.i.d. Bernoulli draws conditional on the seed's calibration set; the paper explicitly notes the alternative hypergeometric model for sampling without replacement from a finite population. If the audit stream shifts relative to deployment, the certificate is invalid, as shown in Table X.
  • domain assumption Each stage satisfies a marginal per-step risk guarantee p_k <= alpha_k via split conformal or other calibration.
    Definition 3 frames the entire problem; the paper instantiates it with split conformal prediction but the framework takes the marginal certificates as given.
  • standard math Standard probabilistic inequalities: Boole's inequality, Hoeffding's inequality, Clopper-Pearson interval validity, Bretagnolle-Huber testing inequality, Hunter's union bound, and association inequalities of Esary-Proschan-Walkup are used as background.
    Invoked in Theorems 1, 8, 9, 10, 11, 12, 17 and the supplementary derivations.
  • domain assumption For Proposition 3 only, stages satisfy the survival-conditional guarantee Pr[E_k | intersection_{j<k} E_j^c] <= alpha_k.
    The paper explicitly states ordinary split CP does not provide this and it is not verified in experiments, so the equal-allocation optimality claim is conditional on this unverified assumption.
  • domain assumption For Theorem 17, the Bernoulli vector Z is associated in the sense of Esary-Proschan-Walkup.
    The duality result requires association, and Remark 6 notes pairwise positive correlation alone is insufficient.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Post-Hoc Trajectory-Risk Certification for Modular LLM-Based Security Agents." pith.science (2026). https://pith.science/paper/47WICBYC

@misc{pith2026260805199,
  author       = {Pith},
  title        = {Pith review of: Post-Hoc Trajectory-Risk Certification for Modular LLM-Based Security Agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/47WICBYC}},
  note         = {Machine review of arXiv:2608.05199}
}
read the original abstract

Autonomous security agents operate as staged pipelines, such as classifying network traffic and then attributing attacks to a specific technique. Split conformal prediction gives each stage finite-sample coverage, but deployment requires a trajectory-level guarantee across the full chain. These guarantees do not compose automatically when stages are independently trained and calibrated. Bonferroni allocation is distribution-free but conservative under correlated errors. We show that a natural pairwise-correlation extension to three or more stages is invalid because it gives a lower rather than an upper bound, and derive a valid spanning-tree alternative. We distinguish whether stages are dependent from whether an audit sample is large enough to certify that dependence, and give matching upper and information-theoretic lower sample-complexity bounds. We also show that coarse-to-fine label selection can create near-perfect measured correlation without learned dependence. On a two-stage intrusion-detection pipeline across 6 open LLMs and 2 datasets, removing this artifact reduces measured correlation from near 1 to 0-0.78. A direct audit of trajectory failure becomes 13.7% tighter than Bonferroni once the audit reaches the required sample size, but is worse when undersized. A modular certificate using per-stage certificates and a pairwise overlap bound yields a positive average gain of 0.6%, quantifying the cost of lacking joint access. Same-model, cross-model, and permuted-pairing tests show that residual dependence reflects shared sample difficulty, not shared model representations. Average trajectory coverage across 12 configurations is 92.7% +/- 2.4% at alpha = 0.10. Under cross-dataset deployment, single-step miscoverage reaches 100% even when accuracy remains 78%, showing that distribution shift destroys calibrated confidence before raw accuracy.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 22 canonical work pages

  1. [1]

    V ovk, A

    V . V ovk, A. Gammerman, and G. Shafer,Algorithmic Learning in a Random World, 01 2005

    work page 2005

  2. [2]

    Inductive confidence machines for regression,

    H. Papadopoulos, K. Proedrou, V . V ovk, and A. Gammerman, “Inductive confidence machines for regression,” inProceedings of the 13th European Conference on Machine Learning, ser. ECML’02. Berlin, Heidelberg: Springer-Verlag, 2002, p. 345–356. [Online]. Available: https://doi.org/10.1007/3-540-36755-1 29

    doi:10.1007/3-540-36755-1 2002

  3. [3]

    On calibration of modern neural networks,

    C. Guo, G. Pleiss, Y . Sun, and K. Q. Weinberger, “On calibration of modern neural networks,”arXiv preprint arXiv:1706.04599, 2017

    arXiv 2017

  4. [4]

    Conformal risk control,

    A. N. Angelopoulos, S. Bates, A. Fisch, L. Lei, and T. Schuster, “Conformal risk control,” inThe Twelfth International Conference on Learning Representations, 2024. [Online]. Available: https: //openreview.net/forum?id=33XGfHLtZg

    2024

  5. [5]

    Rectangular confidence regions for the means of multivariate normal distributions,

    Z. ˇSid´ak, “Rectangular confidence regions for the means of multivariate normal distributions,”Journal of the American Statistical Association, vol. 62, pp. 626–633, 1967

    work page 1967

  6. [6]

    A sharper bonferroni procedure for multiple tests of significance,

    Y . HOCHBERG, “A sharper bonferroni procedure for multiple tests of significance,”Biometrika, vol. 75, no. 4, pp. 800–802, 12 1988. [Online]. Available: https://doi.org/10.1093/biomet/75.4.800

    doi:10.1093/biomet/75.4.800 1988

  7. [7]

    A simple sequentially rejective multiple test procedure,

    S. Holm, “A simple sequentially rejective multiple test procedure,” Scandinavian Journal of Statistics, vol. 6, pp. 65–70, 1979

    1979

  8. [8]

    Controlling the false discovery rate: A practical and powerful approach to multiple testing,

    Y . Benjamini and Y . Hochberg, “Controlling the false discovery rate: A practical and powerful approach to multiple testing,” Journal of the Royal Statistical Society: Series B (Methodological), vol. 57, no. 1, pp. 289–300, 01 1995. [Online]. Available: https://doi.org/10.1111/j.2517-6161.1995.tb02031.x

    arXiv 1995

Show all 35 references
  1. [9]

    Adaptive conformal inference under distribution shift,

    I. Gibbs and E. J. Cand `es, “Adaptive conformal inference under distribution shift,” inProceedings of the 35th International Conference on Neural Information Processing Systems, ser. NIPS ’21. Red Hook, NY , USA: Curran Associates Inc., 2021

    2021

  2. [10]

    Conformal prediction beyond exchangeability,

    R. Barber, E. Cand `es, A. Ramdas, and R. Tibshirani, “Conformal prediction beyond exchangeability,”The Annals of Statistics, vol. 51, 04 2023

    2023

  3. [11]

    Distribution-free, risk-controlling prediction sets,

    S. Bates, A. Angelopoulos, L. Lei, J. Malik, and M. Jordan, “Distribution-free, risk-controlling prediction sets,”J. ACM, vol. 68, no. 6, Sep. 2021. [Online]. Available: https://doi.org/10.1145/3478535

    2021 doi

  4. [12]

    Pasc: Pipeline-aware conformal prediction with joint coverage guarantees for multi-stage NLP and LLM pipelines,

    V . Kotte, “Pasc: Pipeline-aware conformal prediction with joint coverage guarantees for multi-stage NLP and LLM pipelines,”arXiv preprint arXiv:2605.18812, 2026

    2026 arXiv

  5. [13]

    Non-exchangeable conformal risk control,

    A. Farinhas, C. Zerva, D. T. Ulmer, and A. Martins, “Non-exchangeable conformal risk control,” inInternational Conference on Learning Representations (ICLR), 2024

    2024

  6. [14]

    Class-conditional conformal prediction with many classes,

    T. Ding, A. Angelopoulos, S. Bates, M. Jordan, and R. J. Tibshirani, “Class-conditional conformal prediction with many classes,” inAdvances in Neural Information Processing Systems, 2023

    2023

  7. [15]

    Sequential tests of statistical hypotheses,

    A. Wald, “Sequential tests of statistical hypotheses,”Annals of Mathe- matical Statistics, vol. 16, pp. 117–186, 1945

    1945

  8. [16]

    Time-uniform, nonparametric, nonasymptotic confidence sequences,

    S. R. Howard, A. Ramdas, J. McAuliffe, and J. Sekhon, “Time-uniform, nonparametric, nonasymptotic confidence sequences,”Annals of Statistics, vol. 49, no. 2, pp. 1055–1080, 2021

    2021

  9. [17]

    Conformal selective acting: Anytime-valid risk control for RLVR-trained LLMs,

    H. Khosravi and X. Huo, “Conformal selective acting: Anytime-valid risk control for RLVR-trained LLMs,”arXiv preprint arXiv:2605.20270, 2026

    2026 arXiv

  10. [18]

    Audited selective verification for risk-controlled n-1 thermal contingency screening under deployment shift,

    J. Manoharan, “Audited selective verification for risk-controlled n-1 thermal contingency screening under deployment shift,” 2026

    2026

  11. [19]

    Large language models in cybersecurity: State-of-the-art,

    F. N. Motlagh, M. Hajizadeh, M. Majd, P. Najafi, F. Cheng, and C. Meinel, “Large language models in cybersecurity: State-of-the-art,” in International Conference on Information Systems Security and Privacy, 2024

    2024

  12. [20]

    Autoattacker: A large language model guided system to implement automatic cyber-attacks,

    J. Xu, J. W. Stokes, G. McDonald, X. Bai, D. Marshall, S. Wang, A. Swaminathan, and Z. Li, “Autoattacker: A large language model guided system to implement automatic cyber-attacks,”ArXiv, vol. abs/2403.01038, 2024

    2024 arXiv

  13. [21]

    Large language models for cyber security: A systematic literature review,

    H. Xu, S. Wang, N. Li, K. Wang, Y . Zhao, K. Chen, T. Yu, Y . Liu, and H. Wang, “Large language models for cyber security: A systematic literature review,” 2025

    2025

  14. [22]

    Conformal prediction for labelling and updating online models in the presence of concept drift in cybersecurity,

    D. Escudero Garc ´ıa and N. DeCastro-Garc ´ıa, “Conformal prediction for labelling and updating online models in the presence of concept drift in cybersecurity,”Journal of Information Security and Applications, vol. 93, 2025

    2025

  15. [23]

    Inductive conformal anomaly detection for sequential detection of anomalous sub-trajectories,

    R. Laxhammar and G. Falkman, “Inductive conformal anomaly detection for sequential detection of anomalous sub-trajectories,”Annals of Mathematics and Artificial Intelligence, vol. 74, no. 1–2, 2015

    2015

  16. [24]

    Learn then test: Calibrating predictive algorithms to achieve risk control,

    A. N. Angelopoulos, S. Bates, E. J. Cand `es, M. I. Jordan, and L. Lei, “Learn then test: Calibrating predictive algorithms to achieve risk control,” The Annals of Applied Statistics, vol. 19, no. 2, pp. 1641–1662, 2025

    2025

  17. [25]

    An upper bound for the probability of a union,

    D. Hunter, “An upper bound for the probability of a union,”Journal of Applied Probability, vol. 13, no. 3, p. 597–603, 1976

    1976

  18. [26]

    Association of random variables, with applications,

    J. D. Esary, F. Proschan, and D. W. Walkup, “Association of random variables, with applications,”The Annals of Mathematical Statistics, vol. 38, no. 5, pp. 1466–1474, 1967. [Online]. Available: http://www.jstor.org/stable/2238962

    1967

  19. [27]

    T1498: Network denial of service,

    MITRE ATT&CK, “T1498: Network denial of service,” https://attack. mitre.org/techniques/T1498/, 2026

    2026

  20. [28]

    Toward generating a new intrusion detection dataset and intrusion traffic characterization,

    I. Sharafaldin, A. H. Lashkari, and A. A. Ghorbani, “Toward generating a new intrusion detection dataset and intrusion traffic characterization,” in International Conference on Information Systems Security and Privacy, 2018

    2018

  21. [29]

    Quantized autoencoder (qae) intrusion detection system for anomaly detection in resource-constrained iot devices using rt-iot2022 dataset,

    B. S. Sharmila and R. Nagapadma, “Quantized autoencoder (qae) intrusion detection system for anomaly detection in resource-constrained iot devices using rt-iot2022 dataset,”Cybersecurity, vol. 6, pp. 1–15, 2023

    2023

  22. [30]

    Qwen2 technical report,

    A. Yanget al., “Qwen2 technical report,”ArXiv, vol. abs/2407.10671, 2024

    2024 arXiv

  23. [31]

    Gemma 2: Improving open language models at a practical size,

    G. Teamet al., “Gemma 2: Improving open language models at a practical size,”ArXiv, vol. abs/2408.00118, 2024

    2024 arXiv

  24. [32]

    The llama 3 herd of models,

    A. Grattafioriet al., “The llama 3 herd of models,” 2024. [Online]. Available: https://arxiv.org/abs/2407.21783

    2024 arXiv

  25. [33]

    Mistral 7b,

    A. Q. Jianget al., “Mistral 7b,” 2023. [Online]. Available: https://arxiv.org/abs/2310.06825

    2023 arXiv

  26. [34]

    Efficient memory management for large language model serving with pagedattention,

    W. Kwonet al., “Efficient memory management for large language model serving with pagedattention,”Proceedings of the 29th Symposium on Operating Systems Principles, 2023

    2023

  27. [35]

    LoRA: Low-rank adaptation of large language models,

    E. J. Huet al., “LoRA: Low-rank adaptation of large language models,” inInternational Conference on Learning Representations (ICLR), 2022

    2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.