- Rating
- 1
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper presents Mygo, a system for non-interactive Transformer inference under CKKS fully homomorphic encryption. It combines a compact ring configuration, a common post-bootstrap interface allowing four rescale steps, fixed-capacity batching with encrypted metadata to hide occupancy and sequence lengths, and cross-module numerical compilation. The numerical design includes jointly distilled FFN layers with cubic activations and scaling factors folded into affine operations. The evaluation reports plaintext student accuracy on three GLUE tasks, independent encrypted BERT layer tests, complete-graph runtime measurements, resource and numerical ablations, and continuous encrypted execution of LLaMA-3-8B’s first decoder layer. The complete BERT graph processes 128 inputs in 2,788 seconds with 22.37 GiB peak GPU memory on an RTX 5090. However, end-to-end encrypted classification accuracy is not established.
Critical strengths
-
The paper clearly specifies its intended privacy guarantee. The leakage definition and conditional argument distinguish protected occupancy and sequence lengths from observable metadata. This provides useful clarity, although fixed-capacity batching largely follows existing work, and the practical benefit and incremental cost of encrypted metadata remain unestablished.
-
The implementation provides useful feasibility and resource measurements. The paper integrates low-degree model adaptation, site-specific approximations, and affine fusion into a compact FHE configuration. The ring-only control quantifies resource savings, while numerical-alignment ablations and continuous encrypted execution of one LLaMA decoder layer provide supporting evidence. These are useful implementation results, but do not by themselves establish substantial methodological novelty or end-to-end encrypted model accuracy.
Critical weaknesses
-
End-to-end encrypted classification accuracy is not established. Appendix F.1 states that the reported task accuracies are plaintext student results using standard attention, normalization, and pooler operators. The encrypted BERT probes initialize each layer from its reference input, so they do not measure accumulated error through the complete model. Appendix H.2 also acknowledges classification drift in several multi-input runs and limits those results to timing and memory evidence. The complete-graph measurements therefore demonstrate execution, but not the accuracy of the resulting predictions. This is a central validation gap for the proposed encrypted inference system.
-
The co-design offers limited methodological novelty, and its comparative benefit is not established. Fixed-capacity interleaved batching largely follows MOAI. The three elements of the numerical-compilation contribution also provide limited novelty: jointly training low-degree polynomial activations and surrounding affine layers is an established FHE model-adaptation strategy; selecting approximation intervals and circuits for individual sites is an incremental refinement of range-based approximation; and absorbing representation factors into weights and biases is standard affine fusion. The paper does not clearly identify a new algorithm or demonstrate a nontrivial advantage from coordinating these ingredients. Table 2 further combines published timings across different hardware, libraries, capacities, model adaptations, and evaluation scopes. The ring-only control quantifies specific resource savings, but does not establish an end-to-end advantage at comparable accuracy and security.
Questions for authors
-
What accuracy does the full encrypted BERT model achieve on held-out examples with one active input and with all 128 lanes occupied? What causes the classification drift reported in Appendix H.2, and how often do encrypted predictions differ from those of the plaintext student?
-
What evidence shows that Mygo improves runtime and memory use over a baseline at comparable accuracy and security? Can you provide an end-to-end comparison on the same hardware, or ablations that separate the effects of the smaller ring, numerical redesign, and underlying FHE library?
-
Section 3.1 reports a 125.90-bit MATZOV/default estimate and acknowledges that the 128-bit target is not a verified lower bound. What security level is supported for the deployed configuration? What parameter changes would be needed to reach the target under the chosen attack-cost model, and how would they affect runtime and memory?
3 comments
Response to Reviewer 5NBr (Part 2/3, Q1–Q2)
Q1: What accuracy does the full encrypted BERT model achieve on held-out examples with one active input and with all 128 lanes occupied? What causes the classification drift reported in Appendix H.2, and how often do encrypted predictions differ from those of the plaintext student?
For the , Level-4 compact student, all three configurations execute 12 layers, pooler, and classifier.
| Table 13 cohort | FHE label correct | FP32 label correct | Prediction agreement |
|---|---|---|---|
| One unlabeled author example | — | — | 1/1 |
| 127 distinct labeled inputs + author example | 66/127 | 112/127 | 68/128 |
| One labeled example replicated 128 times | 62/128 | 128/128 | 62/128 |
The mixed cohort uses SST-2 validation inputs. On its 127 labeled inputs, 67/127 predictions agree and 60/127 differ; the full-container 68/128 includes the shared unlabeled example. The 2,788 s run is the replicated-lane diagnostic on one RTX 5090. Single/mixed input runs use an RTX PRO 6000.
At the first encoder observation, the shared example's RMSE is 0.04146 with one input and 11,469.4 in the mixed container; aggregate mixed RMSE is 11,348.2. These measurements cover valid tokens in 16 fixed hidden columns. The example's final-logit maximum error is 0.07031/110.615 (single/mixed), while its predicted class is unchanged.
The FP32 student uses standard attention, LayerNorm, and pooler with the same cubic FFNs. Thus the differences include circuit approximation and encrypted execution. Occupancy, neighboring inputs, keys, and encryption randomness vary together: the traces locate divergence by the first layer end without isolating a cause.
Changes: G.2 separates labeled accuracy, agreement, and replicated-lane diagnostics in Table 13 (pp. 37–38, lines 1982–2006). G.3 reports the paired traces, references, and observation scope (p. 38, lines 2008–2020). S1 and the CSV retain per-lane outcomes.
Q2: What evidence shows that Mygo improves runtime and memory use over a baseline at comparable accuracy and security? Can you provide an end-to-end comparison on the same hardware, or ablations that separate the effects of the smaller ring, numerical redesign, and underlying FHE library?
At fixed , Level-4 Boot output, and topology, the compact ring saves 50.3% key payload, 35.1% post-Boot resident memory, and 22.9% synchronized Boot latency. W2's numerical controls hold the ring and online schedule fixed.
The encoder measurements use one RTX PRO 6000, the same Cheddar build, chain, scale , Level-13 Boot output, 256-lane capacity, and 12-layer endpoint. Each row is one run.
| Tables 3/14: light probes | Encoder prefix (s) | Full run (s) | CUDA-pool used peak (GiB) |
|---|---|---|---|
| MYGO teacher | 4,338.28 | 4,376.68 | 37.22 |
| Reference port | 6,945.99 | 6,945.99 | 25.58 |
Prefixes include setup and probes; MYGO's full run also includes its head. MYGO uses its calibrated teacher and 256 active inputs (6,508 tokens); the port uses author weights, fixed shifts, and one five-token example. The 37.5% shorter prefix describes these recorded workloads with their respective checkpoints and active cohorts. See W1 for the accuracy results and Q3 for the security estimates.
For the reference port, Table 14 also records 8,757.00 s for the earlier version. The 20.7% decrease to 6,945.99 s compares versions that differ in both the implementation and the Cheddar build.
Encrypted validity hides occupancy and individual lengths within the public profile and permitted leakage. The 134 ciphertexts comprise 128 full-C pair masks, 1 row mask, 4 Attention length fields, and 1 compatibility shift field (F.3):
| Table 16 configuration | Metadata ciphertexts | Raw RNS payload (GiB) |
|---|---|---|
| Compact student | 134 | 0.262 |
| Level-13 teacher | 134 | 1.243 |
The teacher prepares metadata once per batch: Encode alone takes 6.73 s (0.154% of the light-probe run) within 8.32 s complete preparation, including encryption; online gauge takes 20.24 s.
Changes: Section 6.1/Table 3 identifies the common-backend cohorts (p. 8, lines 412–426); G.4/Table 14 records both port versions and their timing scopes (pp. 38–39, lines 2023–2053). Section 6.2 separates ring and numerical controls (p. 9, lines 434–485); H.1/Table 18A gives paired ring costs (p. 42, lines 2222–2225). G.7/Table 16 lists the metadata inventory and component costs (p. 40, lines 2106–2126; S2–S4).
Response to Reviewer 5NBr (Part 3/3, Q3)
Q3: Section 3.1 reports a 125.90-bit MATZOV/default estimate and acknowledges that the 128-bit target is not a verified lower bound. What security level is supported for the deployed configuration? What parameter changes would be needed to reach the target under the chosen attack-cost model, and how would they affect runtime and memory?
The compact configuration's main-domain minimum is 125.90 bits under the stated MATZOV/GSA model, below the 128-bit target. This estimate concerns the dense main secret (weight 16,384) on the exposed basis, with .
The deployed weight-32 auxiliary Boot secret has a different exposure basis: . On that actual short basis, the evaluated primal BDD hybrid, dual MITM hybrid, and analytical meet-in-the-middle costs are 176.6608, 147.5764, and 210.8826 bits; their minimum is 147.5764 bits. The main-domain four-attack set comprises uSVP, BDD, dual, and dual hybrid. Estimates use , Gaussian error standard deviation 3.2, and the stated balanced-ternary model for random-sign fixed-weight secrets.
A larger-ring candidate keeps the compact full- product, short-basis primes, Level-4 Boot output, and operator topology. At , main-secret weight 32,768 and auxiliary weight 32, its main-domain four-attack minimum is 272.1371 bits. The short-domain dual MITM hybrid estimate is 170.6900 bits.
| Table 18A: fixed- resource control | ||
|---|---|---|
| Synchronized Boot (ms) | 19.278 | 25.018 |
| Evaluation-key payload (GiB) | 8.597 | 17.304 |
| Post-Boot resident GPU memory (GiB) | 15.94 | 24.56 |
These are paired resource microbenchmarks at the compact profile's modulus budget.
A second option is an estimator-only modulus-budget screen. Keeping , main-secret weight 16,384, error standard deviation 3.2, and , a synthetic effective modulus gives a four-attack minimum of 128.0290 bits. This analytical point still requires a compatible RNS chain, Boot precision, and refresh schedule; it has no measured execution cost.
The teacher is specified separately by its Level-13 chain, with and . The reported attack costs apply to the compact chain and its fixed- control.
Changes: Section 3.1 identifies the deployed dense/sparse domains and their estimates (p. 3, lines 146–152). D.2 gives the short-basis attacks, larger-ring candidate, analytical budget screen, and separate teacher-chain binding (pp. 24–25, lines 1290–1329). H.1/Table 18A provides measured resource costs (p. 42, lines 2222–2225; S3).
Response to Reviewer 5NBr (Part 1/3, W1–W2)
The calibrated N=216, Level-13 teacher carries ciphertexts through all 12 layers, pooler, and classifier on the same fixed 256-input SST-2 validation cohort. Both runs achieve 93.36% label accuracy (239/256) and 256/256 agreement with both exact and circuit-matched FP64 references.
Both runs compute all 768 columns; light probes observe four layer-specific columns, full probes all 768. Fitting uses 2,048 training examples; plaintext diagnostics on the fixed validation cohort set positivity margins before FHE.
Thank you for emphasizing the need for final classifier outputs. Table 1's three-task accuracies describe the plaintext compact student with standard attention, LayerNorm, and pooler and the same cubic FFNs. Its independent layer probes start from matching reference inputs. Q1 gives the compact student's continuous results.
Changes:Section 6.1/Table 1 labels plaintext/local results (p. 7, lines 342–361); Table 2/Figure 7 reports teacher classification and coverage (p. 8, lines 378–398). G.2–G.3 specify cohorts and reference models (pp. 37–38, lines 1982–2020; S1).Across the LN1–FFN–LN2 interfaces, MYGO chooses output representations to meet the next module's approximation interval and Boot domain within the shared level budget. The representation factors are distinct from CKKS scale and fold into adjacent affine maps. The compact student's cubic FFN fits four Rescales and places Boot after FC2 contracts 3,072 ciphertexts to 768. SIB combines fixed-capacity interleaving with encrypted validity to keep the public execution graph fixed while protecting occupancy and individual lengths.
Table 19 reports representation-factor controls. The N=215, Level-4 Layer-0 study fixes 100 SST-2 inputs (1,675 valid tokens), parameters, packing, levels, ciphertext counts, and 41 Boot timing events. Each arm uses three fresh processes; boundary entries are medians over valid coordinates.
A separate attention fixture repeats one five-token example across 128 lanes, with three fresh processes per arm (mean ± sample SD). Both selected Table 19A arms use the same degree-31 Mod-8 inverse on [0.1,3.0].
The folded scale change adds no online multiplication, level, or Boot.
Changes:Section 5.1 defines the interface (p. 5, lines 249–259); Section 6.2 explains the bridge/attention controls (p. 9, lines 462–485). H.3 clarifies existing Table 19A/C workloads, matched BASEs, and statistics (pp. 42–44, lines 2256–2326; S2).