Pith. sign in

REVIEW 4 major objections 4 minor 81 references

Design and FPGA Implementation of WOMBAT: A Deep Neural Network Level-1 Trigger System for Jet Substructure Identification and Boosted $H\rightarrow b\bar{b}$ Tagging at the CMS Experiment

T0 review · 4 major / 4 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 An FPGA neural network trigger can select boosted H→bbar jets at a 1 kHz rate with a jet pT threshold about 40 GeV lower than the standard CMS Single Jet 180 trigger.

desk verdict Solid ML trigger proof-of-concept with a real firmware implementation, but the deployability claim is internally contradicted by the 22-cycle latency and the rate normalization needs justification. read the letter →

arxiv 2505.05532 v2 pith:7VVUNCHK submitted 2025-05-08 physics.ins-det hep-ex

Mila Bileska This is my paper · ORCID
classification physics.ins-dethep-ex
keywords Level-1triggerboostedHtobbbarjettaggingFPGAdeepneuralnetworkquantizedCNNknowledgedistillationCMScalorimeter
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 tries to establish that a machine-learned trigger running on the FPGAs of the CMS Level-1 calorimeter trigger can identify Lorentz-boosted H→bbar jets using only trigger-primitive energy grids. The central result is that the WOMBAT Master Model reaches a 1 kHz trigger rate at an offline jet pT of 146.8 GeV, while the standard Single Jet 180 trigger needs 187.4 GeV for the same rate, and the FPGA-synthesizable Apprentice Model reaches 140.4 GeV. If correct, this would let CMS collect boosted Higgs and di-Higgs events in a kinematic window roughly 40 GeV wider than the present baseline, without spending additional trigger bandwidth. The paper also shows the apprentice model fits on a Virtex-7-class FPGA with 22 clock cycles of pre-place-and-route latency, establishing a proof of concept for machine-learned jet tagging at the first trigger level.

What carries the argument

The carrying mechanism is a teacher–student knowledge-distillation pair sharing one input representation: the W-MM uses an Embedded Deterministic Autoencoder (EDA) that treats the azimuthal angle φ as cyclic—circularly padding that dimension while keeping η at full resolution—so the network learns jet substructure that wraps across the detector boundary. The distilled W-AM is an 8-bit quantized CNN with a custom pT-threshold layer that subtracts 30 GeV from each trigger region before convolution, filtering low-energy noise at no latency cost; its fixed two-jet output is what keeps it small enough for the FPGA. JEDI, the comparison baseline, is a fully rule-based pipeline of pileup-subtracted 3×3 energy sums, shape masks, and a 64-element bitonic sorter.

What would settle it

Recompute the 1 kHz rate crossing using the raw event counts without the normalization factor, or split the ZeroBias sample into independent run periods and require the threshold difference to reproduce in each; a movement larger than the ±5.5 GeV bin width would falsify the claim as stated.

Watch

Extended reading notes

Core claim

The claim, stated on the paper's own terms, is that WOMBAT—a convolutional-neural-network trigger built around an Embedded Deterministic Autoencoder and a distilled, 8-bit-quantized apprentice—can localize boosted H→bbar jets directly from a 14×18 grid of calorimeter trigger regions. W-MM achieves the 1 kHz rate at 146.8 GeV and W-AM at 140.4 GeV, compared with 187.4 GeV for Single Jet 180, with W-MM maintaining signal efficiency comparable to the baseline and W-AM accepting a lower-efficiency trade for hardware feasibility. The FPGA implementation of W-AM is reported at 22 clock cycles (137.5 ns) pre-place-and-route with resource usage within the target device, while the rule-based JEDI baseline needs 56 cycles and excessive resources. The paper positions the result as a Phase-1 proof of concept for Phase-2 triggers, where larger latency budgets and finer granularity would allow the same method to be deployed online.

Load-bearing premise

The claimed threshold gain rests on a normalization factor that rescales the WOMBAT and Single Jet 180 rate histograms to a common event count; if that factor is correcting for different event selection rather than plain luminosity, the 40 GeV improvement could be an artifact.

Editorial extensions

If this is right

  • At a fixed 1 kHz per-trigger rate, the ML trigger accepts events with offline jet pT as low as about 147 GeV, opening the 147–187 GeV window that the standard single-jet trigger cannot record.
  • Because W-AM's efficiency is capped by its two-jet output in events with three or more jets, a logical OR of W-AM at low pT with Single Jet 180 at high pT would recover efficiency across the full range without exceeding the rate budget.
  • The 22-cycle pre-place-and-route latency exceeds the Phase-1 14-cycle budget, so the immediate claim is not deployment but feasibility: quantization, thresholding, and pipeline restructuring bring an ML tagger within reach of current FPGA capacity.
  • JEDI's 56-cycle latency and heavy resource use indicate that rule-based boosted-Higgs tagging is the harder path to fit in the Level-1 hardware, strengthening the case for learned features.

Reading between the lines

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

  • Editorial inference: The 40 GeV threshold gain should be read as conditional on the normalization correction in the rate formula; if the raw event counts for the two triggers differ because of acceptance rather than trigger path, the gain could shrink or vanish.
  • Editorial inference: The EDA's circular-φ design is a reusable ingredient for any online trigger that must find objects crossing the detector boundary; one testable extension is to apply the same teacher–student recipe to boosted W/Z or top jets, where the substructure signatures differ but the boundary problem is identical.
  • Editorial inference: A stricter test of the claimed gain would be to run W-AM and Single Jet 180 on the same events with identical n-tuple acceptances, or to measure the rate on a second independent sample; the paper does not report such a cross-check, so the numerical threshold difference is not yet settled.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. This thesis-style manuscript presents WOMBAT, a machine-learning-based Level-1 trigger algorithm for boosted H→bbbar jet identification at CMS, together with its FPGA implementation. WOMBAT comprises a high-performance teacher model (W-MM) and a quantized, FPGA-synthesizable apprentice model (W-AM), and is benchmarked against the existing Single Jet 180 trigger and a hand-crafted rule-based algorithm called JEDI. Rates are evaluated with 2023 CMS ZeroBias data, efficiencies with a dedicated H→bbbar Monte Carlo sample, and firmware cost with HLS synthesis on a Xilinx Virtex-7 device. The headline claims are that W-MM and W-AM sustain a 1 kHz trigger rate at offline jet pT thresholds of 146.8 GeV and 140.4 GeV, respectively, compared with 187.4 GeV for Single Jet 180, and that W-AM fits on the target FPGA with a 22 clock-cycle pre-place-and-route latency, while JEDI needs 56 clock cycles and excessive resources. The manuscript explicitly frames WOMBAT as a proof-of-concept for Phase-2 L1 triggers.

Significance. If the quantitative claims survive scrutiny, the paper would be a useful contribution to the L1 machine-learning trigger literature: it pairs a concrete physics target with real ZeroBias rate evaluation, a complete quantized model description, firmware latency/resource numbers, and links to public repositories, which is a strong level of reproducibility for this type of work. The comparison with a deterministic baseline (JEDI) is also valuable because it demonstrates that simple rule-based designs do not automatically dominate learned triggers. However, the central deployability statement is contradicted by the paper's own latency budget, and the rate comparison depends on an unexplained normalization factor. Those two issues must be resolved before the headline threshold improvements can be taken as a reliable physics result.

major comments (4)
  1. [Ch. V, Sec. 6, Table 7; also abstract] The paper defines the L1T FPGA processing budget as 14 clock cycles (87.5 ns at 6.25 ns/CC) and then states explicitly that none of the algorithms meet that requirement. Table 7 reports W-AM at 22 CC (137.5 ns), i.e., 8 CC over budget, and JEDI at 56 CC. The abstract nonetheless presents the 22 CC latency as evidence that W-AM meets resource constraints, and the concluding summaries describe W-AM as deployable. By the paper's own metric, W-AM does not meet the Phase-1 L1T latency budget. This is an internal contradiction in the central deployability claim, not a matter of external disagreement. Please revise the deployment wording so that W-AM is described as exceeding the Phase-1 latency budget while remaining a hardware-feasibility proof-of-concept, or provide a concrete, reasoned argument for why the 14 CC budget does not apply to this algorithm in its intended position in the L1T chain.
  2. [Ch. IV, Sec. 7, Eq. (73)] Equation (73) multiplies the cumulative event fraction N_{>=pT(i)}/N_total by the factor N_{h0}/N_{h1}, described as a correction for differences in normalization between the WOMBAT and Single Jet 180 rate histograms. If both algorithms process the same ZeroBias sample, the raw event fractions should already be directly comparable; if the n-tuples have different total counts because of different skim conditions or processing paths, the ratio can compensate for those differences but can also imprint them directly on the derived pT thresholds. The manuscript provides no justification that the ratio is unbiased, no test of its stability, and no uncertainty attached to it. Since the headline 40.6 GeV and 47.0 GeV threshold improvements are computed from this formula, this normalization factor is load-bearing. Please either derive the factor from the sample definitions, demonstrate that the conclusion is unchanged on a common subsample, or remove the factor and recompute the rates.
  3. [Abstract and Ch. V, Sec. 3, Table 3] The abstract states that W-MM achieves its lower pT threshold while maintaining comparable signal efficiency to Single Jet 180. Under the standard DeltaR < 0.4 matching used in the analysis, Table 3 reports epsilon(pT) = 0.71 for W-MM versus 0.91 for Single Jet 180 at the 1 kHz threshold. That is a substantial efficiency loss, not a comparable one. Under DeltaR < 0.8 the numbers are closer (0.89 vs 0.95), but the abstract does not state this condition. The text should either qualify the efficiency comparison by the matching criterion or soften the claim. The same precision is needed in the conclusion where W-AM is described as deployable; as noted in the first major comment, that wording is contradicted by the latency table.
  4. [Ch. V, Sec. 6, Table 8] The resource utilization figures are HLS synthesis estimates only, before placement and routing. The manuscript asserts that HLS estimates are usually pessimistic for 7-series devices and therefore adequate for algorithmic comparison, but the central claim that W-AM 'meets resource constraints' on a specific FPGA is stronger than what a pre-P&R report can establish. Please explicitly label the resource numbers as pre-implementation estimates, and either provide post-implementation results or explain in the deployment discussion why the estimates are sufficient for the intended conclusion. This is a correctness-of-wording issue rather than a fundamental flaw, but it feeds the same over-claim identified in the first major comment.
minor comments (4)
  1. [Ch. IV, Sec. 4, after Eq. (70)] The two equations defining the masks r_eta and r_phi are both labeled r_eta; the second line should define r_phi.
  2. [Ch. IV, Sec. 7, Eq. (73)] The conversion factor is written as 40x10^6 / 10^3 [kHz]; the dimensional logic would be clearer if the text stated that it converts a per-bunch-crossing fraction to kHz through the 40 MHz bunch-crossing rate.
  3. [Ch. IV, Sec. 2.2.1, Eq. (27)] The notation in Eq. (27) uses a comma inside the max function arguments in a way that is easy to misread; a clearer formulation would separate the two pooled entries explicitly.
  4. [Appendix F, Table 12] For reproducibility, the repository list would benefit from commit hashes or version tags; the current table identifies projects but not the exact code versions used for the numbers in Tables 7 and 8.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: WOMBAT is benchmarked against external triggers on ZeroBias data and MC, and the flagged latency inconsistency is a correctness problem rather than circularity.

full rationale

The performance chain is empirical rather than derivational. W-MM and W-AM are trained on simulated H to b bbar MC and then evaluated on calorimeter trigger primitive data, with rates measured on 2023 ZeroBias data (0.64 fb-1) and efficiencies measured on H to b bbar MC, benchmarked against Single Jet 180 and JEDI. No fitted parameter is renamed as a prediction, and no load-bearing result is imported from a self-citation; the cited references are external CMS and tool documentation. The Nh0/Nh1 factor in Eq. (73) is a normalization calibration applied to the rate histograms; although the paper does not justify why the raw counts differ, this factor is not constructed from the claimed 1 kHz threshold, so the 40.6 GeV improvement is not forced by construction. The abstract's deployability claim is internally contradicted by Chapter V Section 6, which states: 'As shown in Table 7, none of the algorithms meet the L1T FPGA processing time requirement of < 87.5 ns.' That is a correctness and consistency problem, not a circularity. Likewise, possible overlap between training and evaluation samples is a generalization risk rather than a circular reduction. The derivation is self-contained against external benchmarks, so the circularity score is 0.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The analysis depends on several hand-set thresholds, a pre-calibrated pileup lookup table, a matching radius choice, and an unexplained rate normalization factor. The ML weights are trained rather than derived, and the reported threshold gains are sensitive to these choices. No new physical entities are introduced.

free parameters (6)
  • 30 GeV threshold in W-AM and JEDI = 30 GeV
    Chosen by hand as the pT threshold layer in W-AM and as Ethr in JEDI for pileup suppression; the paper says it was determined to be most effective on average.
  • JEDI pileup offset LUT Delta_i = pre-calibrated 2D lookup table
    Per-pileup-bin and per-energy subtraction values optimized for JEDI; this is a data-derived correction that affects JEDI rates.
  • JEDI veto seed threshold Eseed = 10 GeV
    Fixed parameter in the JEDI veto condition, chosen by hand.
  • JEDI allowed shape masks = five binary masks in Table 1
    Hand-defined topology patterns that determine which 3x3 regions pass as jets; changing them changes JEDI efficiency and rate.
  • Delta R < 0.4 matching criterion = 0.4
    Choice of offline-to-trigger jet matching radius; results change substantially under Delta R < 0.8.
  • Rate normalization factor Nh0/Nh1 = computed from histogram counts
    Equation 73 corrects WOMBAT and Single Jet 180 rate histograms; the paper does not derive or validate this factor, and it directly affects the reported thresholds.
assumptions (6)
  • domain assumption Geant4-based CMS trigger simulation accurately reproduces Level-1 calorimeter trigger primitives.
    All training and efficiency evaluation rely on simulated TPs; if the emulation is biased, the ML predictions and efficiencies inherit the bias.
  • domain assumption 2023 ZeroBias Run2023C data is an unbiased and representative sample of L1T input for rate estimation.
    The rate comparison uses ZeroBias events as a proxy for the full L1T input stream; this is standard but remains an assumption about data quality and pileup conditions.
  • domain assumption Offline AK8 jets with exactly two SoftDrop subjets and one b-hadron per subjet are the correct ground truth for boosted H to bb.
    Efficiency is defined by matching WOMBAT jets to these offline objects; any mis-modeling of b-tagging or SoftDrop changes the efficiency curves.
  • domain assumption Equation 73 is a valid rate conversion formula.
    The rate estimate combines histogram fractions, a normalization correction, and a 40 MHz bunch crossing factor; the normalization correction is not justified.
  • domain assumption The relevant L1T latency budget for this algorithm is 14 clock cycles.
    The paper states this budget in Chapter V, Section 6, and then reports that all implemented algorithms exceed it; the budget choice determines whether W-AM is deployable.
  • domain assumption Training and evaluation MC samples are statistically independent.
    The paper shows separate training and evaluation pT distributions but does not explicitly state that no events were shared between training and the efficiency evaluation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Design and FPGA Implementation of WOMBAT: A Deep Neural Network Level-1 Trigger System for Jet Substructure Identification and Boosted $H\rightarrow b\bar{b}$ Tagging at the CMS Experiment." pith.science (2026). https://pith.science/paper/7VVUNCHK

@misc{pith2026250505532,
  author       = {Pith},
  title        = {Pith review of: Design and FPGA Implementation of WOMBAT: A Deep Neural Network Level-1 Trigger System for Jet Substructure Identification and Boosted $H\rightarrow b\barb$ Tagging at the CMS Experiment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7VVUNCHK}},
  note         = {Machine review of arXiv:2505.05532}
}
abstract

This thesis investigates the physics performance, trigger efficiency, and Field Programmable Gate Array (FPGA) implementation of machine learning (ML)-based algorithms for Lorentz-boosted $H\rightarrow b\bar{b}$ tagging within the CMS Level-1 Trigger (L1T) under Phase-1 conditions. The proposed algorithm, WOMBAT (Wide Object ML Boosted Algorithm Trigger), comprises a high-performance Master Model (W-MM) and a quantized, FPGA-synthesizable Apprentice Model (W-AM), benchmarked against the standard Single Jet 180 and the custom rule-based JEDI (Jet Event Deterministic Identifier) triggers. All algorithms process calorimeter trigger primitive data to localize boosted $H\rightarrow b\bar{b}$ jets. Outputs are post-processed minimally to yield real-valued $(\eta, \phi)$ jet coordinates at trigger tower granularity. Trigger rates are evaluated using 2023 CMS ZeroBias data (0.64 fb$^{-1}$), with efficiency assessed via a Monte Carlo sample of $H\rightarrow b\bar{b}$ offline reconstructed AK8 jets. W-MM achieves a 1 kHz rate at an offline jet $p_T$ threshold of 146.8 GeV, 40.6 GeV lower than Single Jet 180, while maintaining comparable signal efficiency. W-AM reduces the threshold further to 140.4 GeV, with reduced efficiency due to fixed-output constraints and limited multi-jet handling. FPGA implementation targeting the Xilinx Virtex-7 XC7VX690T confirms that W-AM meets resource constraints with a pre-place-and-route latency of 22 clock cycles (137.5 ns). In contrast, JEDI requires excessive resource usage and a 56-cycle latency, surpassing the 14-cycle L1T budget. Originally developed for Run-3 CMS L1T, WOMBAT serves as a proof-of-concept for Phase-2 triggers, where hardware advances will enable online deployment of more sophisticated ML-based L1T systems.

Figures

Figures reproduced from arXiv: 2505.05532 by the authors.

Figure 0.1
Figure 0.1. WOMBAT Logo Design by M. Bileska 5 [PITH_FULL_IMAGE:figures/full_fig_p007_0_1.png] view at source ↗
Figure 1.1
Figure 1.1. Schematic View of The CERN Accelerator Complex and Particle [PITH_FULL_IMAGE:figures/full_fig_p011_1_1.png] view at source ↗
Figure 1.2
Figure 1.2. Luminosity vs. Pileup as Recorded by the CMS Detector During the [PITH_FULL_IMAGE:figures/full_fig_p013_1_2.png] view at source ↗
Figures from the paper (31 more)
Figure 1.3
Figure 1.3. Figure 1.3: Particle Interactions in a Transverse Slice of the CMS Detector [ [PITH_FULL_IMAGE:figures/full_fig_p014_1_3.png]
Figure 1.4
Figure 1.4. Figure 1.4: Dataflow of the L1T During Following Phase-1 Upgrades [ [PITH_FULL_IMAGE:figures/full_fig_p019_1_4.png]
Figure 1.5
Figure 1.5. Figure 1.5: Schematic View of CTP7 and MP7 Cards Constituting the L1 [PITH_FULL_IMAGE:figures/full_fig_p021_1_5.png]
Figure 1.6
Figure 1.6. Figure 1.6: High-Level Diagram of the Phase-2 L1 Trigger Showing Arrows for [PITH_FULL_IMAGE:figures/full_fig_p029_1_6.png]
Figure 2.1
Figure 2.1. Figure 2.1: Standard Model of Particle Physics The Standard Model diagram depicting the bosons (force carrying particles) and the three generations of matter fermions. Possible interactions between species are highlighted, with the mass, charge, and spin shown. Each fermion has …
Figure 2.2
Figure 2.2. Figure 2.2: ggH Production Mechanism of single Higgs and Di-Higgs [PITH_FULL_IMAGE:figures/full_fig_p035_2_2.png]
Figure 2.3
Figure 2.3. Figure 2.3: Di-Higgs Production Processes Through the Gluon-Gluon Fusion [PITH_FULL_IMAGE:figures/full_fig_p037_2_3.png]
Figure 2.4
Figure 2.4. Figure 2.4: Visualization of Particle Decay Collimation With Increasing [PITH_FULL_IMAGE:figures/full_fig_p039_2_4.png]
Figure 2.5
Figure 2.5. Figure 2.5: Phase-2 Physics Reach Based on L1T System [ [PITH_FULL_IMAGE:figures/full_fig_p040_2_5.png]
Figure 3.1
Figure 3.1. Figure 3.1: Phase-1 CMS Calorimeter Trigger Tower Segmentation [PITH_FULL_IMAGE:figures/full_fig_p044_3_1.png]
Figure 3.2
Figure 3.2. Figure 3.2: Raw and Processed Calorimeter TP Display (Event 3468) [PITH_FULL_IMAGE:figures/full_fig_p045_3_2.png]
Figure 4.1
Figure 4.1. Figure 4.1: Cumulative Distribution Function Comparison for W-AM With and [PITH_FULL_IMAGE:figures/full_fig_p057_4_1.png]
Figure 4.2
Figure 4.2. Figure 4.2: Cumulative Distribution Function Comparison for W-AM With and [PITH_FULL_IMAGE:figures/full_fig_p059_4_2.png]
Figure 4.3
Figure 4.3. Figure 4.3: Cumulative Distribution Function Comparison of W-MM and W-AM [PITH_FULL_IMAGE:figures/full_fig_p060_4_3.png]
Figure 4.4
Figure 4.4. Figure 4.4: η and ϕ Prediction Distributions Compared To Ground Truth for W-MM and W-AM shown in [PITH_FULL_IMAGE:figures/full_fig_p061_4_4.png]
Figure 4.5
Figure 4.5. Figure 4.5: Raw Prediction Spray on η − ϕ Grid for W-MM and W-AM ceed the constraints of the target FPGA, rendering it unsuitable for deployment. In contrast, W-AM represents the highest-performing model that meets the hardware limitations, making it the most viable option for F…
Figure 5.1
Figure 5.1. Figure 5.1: W-MM Good Match TP Display - Event 2687 [PITH_FULL_IMAGE:figures/full_fig_p077_5_1.png]
Figure 5.3
Figure 5.3. Figure 5.3: W-MM Jet Multiplicity Mismatch TP Display - Event 689 [PITH_FULL_IMAGE:figures/full_fig_p077_5_3.png]
Figure 5.5
Figure 5.5. Figure 5.5: W-AM Good Match TP Display - Event 3360 [PITH_FULL_IMAGE:figures/full_fig_p078_5_5.png]
Figure 5.7
Figure 5.7. Figure 5.7: W-AM Jet Multiplicity Mismatch TP Display - Event 830 [PITH_FULL_IMAGE:figures/full_fig_p079_5_7.png]
Figure 5.9
Figure 5.9. Figure 5.9: W-MM and Single Jet 180 Trigger Rate vs. Offline pT With R(pT ) = 1 kHz Threshold [PITH_FULL_IMAGE:figures/full_fig_p081_5_9.png]
Figure 5.11
Figure 5.11. Figure 5.11: ∆R Matching Condition Visualization for ∆R Separations of 0.80, 0.40, and 0.02 As shown in [PITH_FULL_IMAGE:figures/full_fig_p083_5_11.png]
Figure 5.12
Figure 5.12. Figure 5.12: W-MM and Single Jet 180 Trigger Efficiency vs. Offline pT With ϵ(pT ) Threshold for ∆R < 0.4 [PITH_FULL_IMAGE:figures/full_fig_p084_5_12.png]
Figure 5.14
Figure 5.14. Figure 5.14: W-MM and Single Jet 180 Trigger Efficiency vs. Offline pT With ϵ(pT ) Threshold for ∆R < 0.8 [PITH_FULL_IMAGE:figures/full_fig_p084_5_14.png]
Figure 5.16
Figure 5.16. Figure 5.16: MC Evaluation Dataset Jet Multiplicity per Event Leading Order (LO) [PITH_FULL_IMAGE:figures/full_fig_p085_5_16.png]
Figure 5.17
Figure 5.17. Figure 5.17: Efficiency Curve of W-AM and Single Jet 180 Compared to the Maximal [PITH_FULL_IMAGE:figures/full_fig_p086_5_17.png]
Figure 5.18
Figure 5.18. Figure 5.18: Training H → b ¯b MC Dataset Jet pT Distribution When the matching criterion is relaxed from ∆R < 0.4 to ∆R < 0.8, the W-AM model exhibits a similar overall efficiency trend, but with higher ϵ(pT ) values, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p087_5_…
Figure 5.19
Figure 5.19. Figure 5.19: W-MM Trigger Efficiency vs. Offline pT on Events with Jet Multiplicity of 2 [PITH_FULL_IMAGE:figures/full_fig_p089_5_19.png]
Figure 5.21
Figure 5.21. Figure 5.21: W-MM Trigger Efficiency vs. Offline pT on Events with Jet Multiplicity of 2 for ∆R < 0.8 [PITH_FULL_IMAGE:figures/full_fig_p089_5_21.png]
Figure 5.23
Figure 5.23. Figure 5.23: Rate vs Offline pT for W-AM, JEDI, and Single Jet 180 With Threshold At R(pT ) = 1 kHz A contributing factor to JEDI’s reduced efficiency is the super-region activity veto condition, as detailed in Chapter IV, Section 4, and summarized in [PITH_FULL_IMAGE:figures/f…
Figure 5.25
Figure 5.25. Figure 5.25: Trigger Efficiency vs. Offline pT for W-AM, JEDI, and Single Jet 180 Evaluated on Jet Multiplicity of 2 Events (∆R < 0.4) using a fixed 3 × 3 grid structure, computing energy sums, and applying predefined veto conditions on a per-candidate basis. As a result, the al…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

81 extracted references · 52 canonical work pages

  1. [1]

    CERN Collaboration. (n.d.). The History of CERN . Timeline CERN. https://timeline.web.cern.ch/timeline-header/89

  2. [2]

    Fartoukh, S., et al. (2021). LHC Configuration and Operational Scenario for Run 3. CERN Docu- ment Server. CERN-ACC-2021-0007

    work page 2021

  3. [3]

    Ciesla, K. (2024). Heavy-ion physics at ATLAS and CMS. CERN Document Server

    work page 2024

  4. [4]

    Vretenar, M., et al. (2008). The LINAC4 Project: Overview and Status . CERN Document Server

    work page 2008

  5. [5]

    Reich, K. H. (1969). The CERN Proton Synchrotron Booster. IEEE Transactions on Nuclear Sci- ence, 16(3), 959-961. https://doi.org/10.1109/tns.1969.4325414

    work page arXiv 1969

  6. [6]

    Albright, S., et al. (2021). New Longitudinal Beam Production Methods In The CERN Proton Syn- chrotron Booster. JACoW Publishing. https://doi.org/10.18429/JACoW-IPAC2021-THPAB183

    work page doi:10.18429/jacow-ipac2021-thpab183 2021

  7. [7]

    CERN Collaboration. (1959). Proton Synchrotron . CERN Document Server

    work page 1959

  8. [8]

    Hori, M., & Walz, J. (2013). Physics at CERN’s Antiproton Decelerator. Progress in Particle and Nuclear Physics, 72, 206-253. https://doi.org/10.1016/j.ppnp.2013.02.004

    work page doi:10.1016/j.ppnp.2013.02.004 2013

Show all 81 references
  1. [9]

    Kugler, E. (1993). The Isolde Facility at the CERN PS booster. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms , 79(1-4), 322-325. https://doi.org/10.1016/0168-583x(93)95355-9

    1993 doi

  2. [10]

    Etisken, O. (2023). CERN Super Proton Synchrotron and an Alternative Design as Prebooster Ring for the Future Circular Collider e + e - Injector complex. Physical Review Accelerators and Beams. https://doi.org/10.1103/PhysRevAccelBeams.26.081601

    2023 doi

  3. [11]

    Podlaski, P. (2024). NA61/SHINE Overview. arXiv. https://doi.org/arXiv:2402.10973

    2024 arXiv

  4. [12]

    Banerjee, D., et al. (2021). The North Experimental Area at the Cern Super Proton Synchrotron. CERN Document Server. CERN-ACC-NOTE-2021-0015

    2021

  5. [13]

    (n.d.-a)

    CERN Collaboration. (n.d.-a). CERN’s Accelerator Complex. CERN. https://www.home.cern/science/accelerators/accelerator-complex

  6. [14]

    CMS Collaboration. (2006). CMS Physics: Technical Design Report Volume 1: Detector Perfor- mance and Software. https://doi.org/10.2172/2510878

    2006 doi

  7. [15]

    ATLAS Collaboration. (1999). ATLAS Detector and Physics Performance Volume I. CERN/LHCC 99-14

    1999

  8. [16]

    Crosetto, D. B. (1998). A Large Hadron Collider Beauty Experiment for Precision Measurements of CP Violation and Rare Decays. LHCB Technical Proposal. https://doi.org/10.2172/762189

    1998 doi

  9. [17]

    Luciano, M., & Vito, M. (2012). Conceptual Design Report for the Upgrade of the ALICE ITS . CERN-LHCC-2012-005

    2012

  10. [18]

    ATLAS Collaboration. (2016). Measurement of the higgs boson production cross section in Hadronic Final States in 13 TEV proton-proton collisions with the Atlas Detector. Physical Re- view Letters. https://doi.org/10.1103/PhysRevLett.117.182002 114 M. Bileska REFERENCES

    2016 doi

  11. [19]

    Morovic, S. (2023). CMS detector: Run 3 status and plans for Phase. International Workshop on Deep-Inelastic Scattering and Related Subjects. https://doi.org/arXiv:2309.02256

    2023 arXiv

  12. [20]

    CMS Collaboration. (n.d.). CMS luminosity results from pub- lic documents. LumiPublicResultsPAS &lt; CMSPublic &lt; TWiki. https://twiki.cern.ch/twiki/bin/view/CMSPublic/LumiPublicResultsPAS

  13. [21]

    CMS Collaboration. (2008). The CMS experiment at the CERN LHC. Journal of Instrumentation. https://doi.org/10.1088/1748-0221/3/08/S08004

    2008 doi

  14. [22]

    CMS Collaboration. (2017). Particle-flow reconstruction and global event description with the CMS detector. Journal of Instrumentation. https://doi.org/10.1088/1748-0221/12/10/P10003

    2017 doi

  15. [23]

    Dordevic, M. (2018). The CMS Particle Flow Algorithm. EPJ Web of Conferences. https://doi.org/10.1051/epjconf/201819102016

    2018

  16. [24]

    Bols, E., et al. (2020). Jet flavour classification using DeepJet. Journal of Instrumentation, 15(12). https://doi.org/10.1088/1748-0221/15/12/p12012

    2020 doi

  17. [25]

    CMS Collaboration. (2017b). The CMS trigger system. Journal of Instrumentation . https://doi.org/10.1088/1748-0221/12/01/P01020

    2017 doi

  18. [26]

    Development of the CMS detector for the CERN LHC Run 3

    CMS Collaboration. “Development of the CMS detector for the CERN LHC Run 3.” Journal of Instrumentation, 2024, https://doi.org/10.1088/1748-0221/19/05/P05064

    2024 doi

  19. [27]

    LHC Collaboration. (2024). Longer term LHC schedule . LHC Commisioning. https://lhc- commissioning.web.cern.ch/schedule/LHC-long-term.htm

    2024

  20. [28]

    CMS Collaboration. (2021). The Phase-2 Upgrade of the CMS Data Acquisition and High Level Trigger Technical Design Report. CERN Document Server. https://doi.org/CERN-LHCC-2021-007

    2021

  21. [29]

    Tuura, L., Meyer, A., Segoni, I., & Ricca, G. D. (2010). CMS Data Quality Monitoring: Systems and experiences. Journal of Physics: Conference Series, 219(7) , 072020. https://doi.org/10.1088/1742- 6596/219/7/072020

    2010 doi

  22. [30]

    CMS Collaboration. (2013). CMS Technical Design Report for the Level-1 Trigger Upgrade.CERN Document Server. CMS-TDR-012

    2013

  23. [31]

    CMS Collaboration. (2000). The TriDAS Project Technical Design Report, Volume 1: The Trigger Systems. CERN Document Server. CERN/LHCC 2000-38

    2000

  24. [32]

    Tapper, A. (2017). The CMS level-1 trigger for LHC run II. Proceedings of 38th International Conference on High Energy Physics — PoS(ICHEP2016), 242. https://doi.org/10.22323/1.282.0242

    2017 doi

  25. [33]

    Zabi, A. (2015). The CMS calorimeter trigger upgrade for the LHC run II. Proceed- ings of Technology and Instrumentation in Particle Physics 2014 — PoS(TIPP2014), 414. https://doi.org/10.22323/1.213.0414

    2015 doi

  26. [34]

    CMS Collaboration. (2013a). CMS level-1 upgrade calorimeter trigger prototype development. Journal of Instrumentation. https://doi.org/10.1088/1748-0221/8/02/C02013

    2013 doi

  27. [35]

    CMS Collaboration. (2020). Performance of the CMS Level-1 trigger in proton-proton collisions at√s = 13 TeV.Journal of Instrumentation. https://doi.org/10.1088/1748-0221/15/10/P10017 115 M. Bileska REFERENCES

    2020 doi

  28. [36]

    (n.d.-b)

    CMS Collaboration. (n.d.-b). High level trigger. SWGuideHighLevelTrigger &lt; CMSPublic &lt; TWiki. https://twiki.cern.ch/twiki/bin/view/CMSPublic/SWGuideHighLevelTrigger

  29. [37]

    CERN Collaboration. (2020). High-Luminosity Large Hadron Collider (HL-LHC) Technical design report. CERN Document Server. CERN-2020-010

    2020

  30. [38]

    Pasztor, G. (2022). The Phase-2 Upgrade of the CMS Detector. Proceedings of Science

    2022

  31. [39]

    CMS Collaboration. (2017c). The Phase-2 Upgrade of the CMS Level-1 Trigger. CERN Document Server. CMS-TDR-017

    2017

  32. [40]

    Claude, D., & Ball, A. (2015). Technical Proposal for the Phase-II Upgrade of the CMS Detector. CERN Document Server. CMS-TDR-15-02

    2015

  33. [41]

    Thomson, M. (2021). Modern Particle Physics. Cambridge University Press

    2021

  34. [42]

    Xing, Z. (2014). Neutrino Physics. CERN Document Server. https://doi.org/arXiv:1406.7739

    2014 arXiv

  35. [43]

    Tully, C. G. (2013). Elementary Particle Physics in a nutshell. World Publishing Corporation

    2013

  36. [44]

    Glashow, S. L. (1961). Partial-symmetries of weak interactions. Nuclear Physics, 22(4), 579-588. https://doi.org/10.1016/0029-5582(61)90469-2

    1961 doi

  37. [45]

    CMS Collaboration. (2025). Constraints on the Higgs boson self-coupling from the combination of single and double Higgs boson production in proton-proton collisions at √s = 13 TeV. Physics Letters B. https://doi.org/10.1016/j.physletb.2024.139210

    2025

  38. [46]

    CMS Collaboration. (2012). Observation of a new boson at the LHC with the CMS experiment. Physics Letters B. https://doi.org/10.1016/j.physletb.2012.08.021

    2012 doi

  39. [47]

    ATLAS Collaboration. (2012). Observation of a new particle in the search for the Stan- dard Model Higgs boson with the ATLAS detector at the LHC. Physics Letters B. https://doi.org/10.1016/j.physletb.2012.08.020

    2012 doi

  40. [48]

    Physical Measurement Laboratory

    Reduced Planck constant in eV s . Physical Measurement Laboratory. (n.d.). https://physics.nist.gov/cgi-bin/cuu/Value?hbarev

  41. [49]

    2024: Summary tables.Particle Data Group. (n.d.). https://pdg.lbl.gov/2024/tables/contents_tables.html

    2024

  42. [50]

    (n.d.-a)

    CMS Collaboration. (n.d.-a). CERN accelerating science. Life of the Higgs boson | CMS Experi- ment. https://cms.cern/news/life-higgs-boson

  43. [51]

    ATLAS Collaboration. (2018). Observation of H- >bb decays and VH production with the ATLAS detector. Physics Letters B. https://doi.org/10.1016/j.physletb.2018.09.013

    2018 doi

  44. [52]

    CMS Collaboration. (2020a). Inclusive search for highly boosted Higgs bosons decaying to bottom quark-antiquark pairs in proton-proton collisions at√s = 13 TeV. Journal of High Energy Physics. https://doi.org/10.1007/JHEP12(2020)085

    2020 doi

  45. [53]

    CMS Collaboration. (2019). Measurement and interpretation of differential cross sections for Higgs boson production at √s = 13 TeV. Physics Letters B . https://doi.org/10.1016/j.physletb.2019.03.059

    2019 doi

  46. [54]

    M., & Zeng, M

    Dawson, S., Lewis, I. M., & Zeng, M. (2015). Usefulness of effective field theory for boosted higgs production. Physical Review D, 91(7). https://doi.org/10.1103/physrevd.91.074012 116 M. Bileska REFERENCES

    2015 doi

  47. [55]

    Grazzini, M., Ilnicka, A., Spira, M., & Wiesemann, M. (2017). Effective field theory in quest to Parametrise higgs properties: The Transverse Momentum Spectrum Case. Journal of Physics: Conference Series, 873, 012050. https://doi.org/10.1088/1742-6596/873/1/012050

    2017 doi

  48. [56]

    Li, Y.-Y., Nicolaidou, R., & Paganis, S. (2019a). Exclusion of heavy, broad resonances from precise measurements of WZ and VH Final States at the LHC. The European Physical Journal C , 79(4). https://doi.org/10.1140/epjc/s10052-019-6858-5

    2019 doi

  49. [57]

    F., & Re, E

    Bishara, F., Haisch, U., Monni, P. F., & Re, E. (2017). Constraining light- quark Yukawa couplings from higgs distributions. Physical Review Letters, 118(12). https://doi.org/10.1103/physrevlett.118.121801

    2017 doi

  50. [58]

    Grazzini, M., Ilnicka, A., Spira, M., & Wiesemann, M. (2017b). Modeling BSM effects on the higgs transverse-momentum spectrum in an EFT approach. Journal of High Energy Physics, 2017(3). https://doi.org/10.1007/jhep03(2017)115

    2017 doi

  51. [59]

    CMS Collaboration. (2013c). Identification of b-quark jets with the CMS experiment. Journal of Instrumentation. https://doi.org/10.1088/1748-0221/8/04/p04013

    2013 doi

  52. [60]

    Coleman, E., et al. (2018). The importance of calorimetry for highly-boosted jet substructure. Journal of Instrumentation, 13(01). https://doi.org/10.1088/1748-0221/13/01/t01003

    2018 doi

  53. [61]

    Guest, D., Cranmer, K., & Whiteson, D. (2018). Deep learning and its application to LHC physics. Annual Review of Nuclear and Particle Science , 68(1), 161-181. https://doi.org/10.1146/annurev- nucl-101917-021019

    2018 doi

  54. [62]

    Zabi, A., et al. (2020). The Phase-2 Upgrade of the CMS Level-1 Trigger. CERN Document Server. CERN-LHCC-2020-004

    2020

  55. [63]

    Alwall, J., et al. (2014). The automated computation of tree-level and next-to-leading order dif- ferential cross sections, and their matching to Parton shower simulations. Journal of High Energy Physics, 2014(7). https://doi.org/10.1007/jhep07(2014)079

    2014 doi

  56. [64]

    Lavesson, N., & Lönnblad, L. (2008). Merging Parton showers and matrix elements—back to basics. Journal of High Energy Physics, 2008(04), 085-085. https://doi.org/10.1088/1126- 6708/2008/04/085

    2008 doi

  57. [65]

    Bierlich, C., et al. (2022). A comprehensive guide to the physics and usage of Pythia 8.3. SciPost Physics Codebases. https://doi.org/10.21468/scipostphyscodeb.8

    2022 doi

  58. [66]

    GEANT4 Collaboration. (2002). G eant4 - A Simulation Toolkit. https://doi.org/SLAC-PUB-935

    2002

  59. [67]

    (n.d.-a)

    CMS Collaboration. (n.d.-a). CMS guide on how to calculate luminosity.CMS Guide on how to cal- culate luminosity | CERN Open Data Portal. https://opendata.cern.ch/docs/cms-guide-luminosity- calculation

  60. [68]

    Scripy. PyPI. (n.d.). https://pypi.org/project/Scripy/

  61. [69]

    S., & Pitts, W

    McCulloch, W. S., & Pitts, W. (1943). A logical calculus of the ideas immanent in nervous activity. The Bulletin of Mathematical Biophysics, 5(4), 115-133. https://doi.org/10.1007/bf02478259

    1943 doi

  62. [70]

    Norrstig, A. (2019). Visual Object Detection using Convolutional Neural Networks in a Virtual Environment. Department of Electrical Engineering Linkoping University. 117 M. Bileska REFERENCES

    2019

  63. [71]

    (n.d.-a)

    QKeras Collaboration. (n.d.-a). QKeras: A quantization deep learning library for Tensorflow Keras. GitHub. https://github.com/google/qkeras

  64. [72]

    HLS4ML Collaboration. (n.d.). hls4ml 1.1.0 documentation. https://fastmachinelearning.org/hls4ml/

  65. [73]

    TensorFlow

    Tensorflow. TensorFlow. (n.d.). https://www.tensorflow.org/

  66. [74]

    TensorFlow

    Tf.Keras.Layers.Lambda. TensorFlow. (n.d.). https://www.tensorflow.org/api_docs/python/tf /k- eras/layers/Lambda

  67. [75]

    Xilinx Sales | FPGAs (Field Programmable Gate Array) | CPLDs (Complex Programmable Logic Devices)

    Xilinx: Fpgas: Cplds. Xilinx Sales | FPGAs (Field Programmable Gate Array) | CPLDs (Complex Programmable Logic Devices). (n.d.). https://www.xilinxsemi.com/?search=XC7VX690T-2FFG1927I

  68. [76]

    Fingeroff, M. (2010). High-level synthesis: Blue Book. Xlibris Corporation: Mentor Graphics Cor- poration

    2010

  69. [77]

    Perry, D. L. (2002). VHDL: Programming by example. McGraw-Hill

    2002

  70. [78]

    Duarte, J., et al. (2018). Fast inference of deep neural networks in fpgas for particle physics. Journal of Instrumentation, 13(07). https://doi.org/10.1088/1748-0221/13/07/p07027

    2018 doi

  71. [79]

    Goswami, P., & Bhatia, D. (2022). Predicting post-route quality of results estimates for HLS designs using machine learning. 2022 23rd International Symposium on Quality Electronic Design (ISQED), 45-50. https://doi.org/10.1109/isqed54688.2022.9806201

    2022

  72. [80]

    Staff. (n.d.). Field-programmable gate arrays explained: A high-level introduction to FPGAs. https://files.digilent.com/reference/Field_Programmable_Gate_Arrays_Explained.pdf

  73. [81]

    Neutelings, I. (2025). CMS Coordinate System. TikZ.net. https://tikz.net /axis3d_cms/ 118

    2025

Pith tools

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