- Rating
- 4(+1)
- Confidence
- 3
- Informativeness
- 3
- Presentation
- 3
- Soundness
- 3
Read review
Paper summary
This paper studies whether test-time adaptation (TTA) can recover utility lost when classifiers are trained under label differential privacy (LabelDP), without consuming additional privacy budget for the protected training labels. The paper combines existing LabelDP mechanisms, primarily LP-MST, with source-free TTA methods including batch-normalization adaptation and TENT.
On the theoretical side, the authors show that symmetric randomized response contracts the class posterior toward the uniform distribution while preserving the Bayes argmax, and that its population cross-entropy objective can be written as a mixture of the clean objective and a uniform-label objective. Under additional PL, quadratic-growth, and regularity assumptions, they derive bounds on clean-risk excess and changes in population BN statistics. They separately analyze convergence of entropy minimization under a fixed TENT objective. The paper carefully notes that these results do not directly guarantee that TTA improves classification accuracy.
Empirically, the paper reports that TENT improves clean CIFAR-10 performance at every evaluated finite privacy budget, with the improvement increasing from 0.87 percentage points at to 3.62 points at , while degrading a matched non-private model by 0.79 points. On corrupted data, TTA produces much larger gains, although the non-private model benefits at least as much, and the experiments suggest that most of this robustness gain comes from recomputing target-batch BN statistics rather than entropy minimization itself.
The paper further studies source-model quality, batch size, stream composition, calibration, prediction flips, pretrained representations, and several TTA variants. Notably, TTA fails when the source model is near chance, with very small batches, or under strongly label-correlated test batches.
Critical strengths
- The problem is interesting and practically relevant. Connecting LabelDP with deployment-time adaptation is a natural but underexplored question. The post-processing argument is clean: under the stated threat model, adaptation using only the released model and unlabeled inputs does not spend additional privacy budget for the original training labels.
- The empirical characterization is considerably stronger than a simple “TENT improves accuracy” result. I particularly appreciate the investigation of batch size, stream composition, source quality, arrival-order dependence, calibration, and individual prediction flips. The negative results are informative. For example, at , a label-correlated stream increases error from 23.86% for source inference to 62.96% for BN adaptation and 66.07% for TENT.
- The main CIFAR-10 LabelDP results have reasonable seed analysis. The five-seed results support the main TENT improvements; e.g., clean errors at move from to . Corruption gains are also much larger than the reported variability.
- The paper is unusually explicit about its limitations. In particular, it acknowledges the gap between symmetric-RR theory and LP-MST experiments, fixed-objective gradient-descent theory and streaming Adam-based TENT, and population BN statistics versus the actual target-statistics mismatch.
Critical weaknesses
See questions.
Questions for authors
- Could the authors provide a matched softened-supervision control? In particular, I would find a non-private model trained with label smoothing matched to , ideally with the same mixup configuration and seeds, highly informative. If TENT exhibits the same clean-data improvement there, how would the authors revise the claimed mechanism?
- Could the non-private baseline be repeated across multiple seeds and reported as paired source-to-TTA changes? This would strengthen the central observation that TTA changes from beneficial under LabelDP to harmful without privatization.
- For Theorem 5, what guarantees existence of the finite entropy minimizer ? If this is not guaranteed for the parameterization considered, could the statement be reformulated in terms of an infimum or an explicitly restricted parameter domain?
3 comments
Thanks for the reply. I have raised the score to support this paper acceptance.
Thank you for raising your score and for your helpful feedback! We will include the promised revisions in the final version.
Thank you for the careful review and the precise questions. The control the reviewer proposed turned out to be very informative. We show that matched label smoothing does not achieve the same clean data gains, which points to label randomization as the source for the gains.
Q1. Matched softened supervision with Label Smoothing
We trained label-smoothing controls using clean labels smoothed by 1−ρ(ε) = 0.853, 0.610 and 0.157 (ε = 1, 2, 4), with the same LP-MST pipeline and seeds as the private models and α=4 mixup. At ε=8 the coefficient is 0.003, so we use the non-private model with the same α=4 mixup as the ε=8 private model.
The label-smoothing controls on their own. CIFAR-10 test error (%) over 5 seeds. Change is TENT minus source, paired per seed:
Under the corruption dataset, they benefit like every other model. On clean data, TENT does not show the gain it gives the matched LabelDP models:
To answer your question, since TTA does not reproduce the gain on label smoothing, the mechanism stands, and the control makes the argument sharper. Lemma 2's expected target is what private training aims at. Training on that target alone costs no accuracy and leaves nothing to recover. The recoverable loss comes from randomized response fixing one noisy label per example. Training on those labels leaves BN statistics that do not match clean inputs, and an underconfident but usable ranking. TTA fixes both, and the extra gain decreases as the amount of label randomization decreases, from 4.55 points at ε=1 near to zero at ε=8. Proposition 1 and Lemma 2 explain why the private target is underconfident and why its ranking is worth recovering.
Q2. Multi-seed control and paired changes
Five seeds each, paired per seed, test error:
Severity 5:
Without mixup, as in the submission, TTA raises the clean error of the non-private control. With α=4 mixup, we observe a small gain for TTA on CIFAR-10. Our revision compares the LabelDP gains with this non-private control and with the label-smoothing control.
Table 1 and Table 15 (formerly Table 10) now report five-seed means: at ε=2,4,8, clean error goes from 15.95, 8.94 and 7.28 before TENT to 13.18, 7.50 and 6.52 after. The paired gains agree with our original seed within one standard deviation.
Q3. Theorem 5
Thank you for catching that. We restate the theorem with H∗:=infψH(ψ) over the stated domain and Δt:=H(ψt)−H∗. The proof only uses H(ψ†) as a lower bound, so it goes through unchanged:
H(ψT)−H∗≤(1−LHμH)T[H(ψ0)−H∗].
The same assumptions then give the minimizer back. Each step has length ∥ψt+1−ψt∥=∥∇H(ψt)∥/LH≤2Δt/LH, which decays geometrically. So the path has finite length, and the iterates converge to some ψ∞ with H(ψ∞)=H∗. Defining ψ†:=ψ∞ keeps Corollary 6 valid as stated.
The PL condition cannot hold along a path that drifts toward infinite scale, corresponding to the confidently-wrong collapse that can occur under entropy minimization.