Provable Joint Decontamination for Benchmarking Multiple Large Language Models

Benchmark data contamination has become a central challenge in LLM evaluation: when evaluation examples appear in the training data of one or more audited models, reported performance can be inflated and cross-model comparisons become unreliable. A broad line of training-data detection work designs scores to quantify how strongly a model memorizes a given data point, but these score-based methods lack theoretical guarantees. Recent conformal approaches provide provable false-identification control for a single model; however, applying them separately to each model can produce model-specific benchmarks, undermining fair comparison across models. In this work, we formalize multi-model benchmark decontamination as a joint selection problem and propose Joint Envelope Conformal Selection (JECS), a conformal procedure that enables global contamination rate (GCR) control under stated assumptions. Specifically, JECS computes per-model conformal p-values, aggregates them by the per-item maximum, and reconstructs a conservative envelope of the max-p null distribution from right-tail observations above a data-driven threshold. By applying the adaptive Benjamini-Hochberg (BH) procedure to the envelope-rescaled values, we select a benchmark with provable GCR control. Extensive experiments across various models and benchmarks demonstrate that JECS achieves higher power than the max-p baseline while consistently maintaining the target GCR control.

Efficient Membership Inference Attacks by Bayesian Neural Network

Membership Inference Attacks (MIAs) aim to estimate whether a specific data point was used in the training of a given model. Previous attacks often utilize multiple reference models to approximate the conditional score distribution, leading to significant computational overhead. While recent work leverages quantile regression to estimate conditional thresholds, it fails to capture epistemic uncertainty, resulting in bias in low-density regions. In this work, we propose a novel approach - Bayesian Membership Inference Attack (BMIA), which performs conditional attack through Bayesian inference. In particular, we transform a trained reference model into Bayesian neural networks by Laplace approximation, enabling the direct estimation of the conditional score distribution by probabilistic model parameters. Our method addresses both epistemic and aleatoric uncertainty with only a reference model, enabling efficient and powerful MIA. Extensive experiments on five datasets demonstrate the effectiveness and efficiency of BMIA.

A Modulo Sampling Hardware Prototype and Reconstruction Algorithm Evaluation

Analog-to-digital converters (ADCs) play a vital important role in any devices via manipulating analog signals in a digital manner. Given that the amplitude of the signal exceeds the dynamic range of the ADCs, clipping occurs and the quality of the digitized signal degrades significantly. In this paper, we design a joint modulo sampling hardware and processing prototype which improves the ADCs' dynamic range by folding the signal before sampling. Both the detailed design of the hardware and the recovery results of various state-of-the-art processing algorithms including our proposed unlimited sampling line spectral estimation (USLSE) algorithm are presented. Additionally, key issues that arise during implementation are also addressed. It is demonstrated that the USLSE algorithm successfully recovers the original signal with a frequency of 2.5 kHz and an amplitude 10 times the ADC's dynamic range, and the linear prediction (LP) algorithm successfully recovers the original signal with a frequency of 3.5 kHz and an amplitude 10 times the ADC's dynamic range.

Exploring Learning Complexity for Downstream Data Pruning

The over-parameterized pre-trained models pose a great challenge to fine-tuning with limited computation resources. An intuitive solution is to prune the less informative samples from the fine-tuning dataset. A series of training-based scoring functions are proposed to quantify the informativeness of the data subset but the pruning cost becomes non-negligible due to the heavy parameter updating. For efficient pruning, it is viable to adapt the similarity scoring function of geometric-based methods from training-based to training-free. However, we empirically show that such adaption distorts the original pruning and results in inferior performance on the downstream tasks. In this paper, we propose to treat the learning complexity (LC) as the scoring function for classification and regression tasks. Specifically, the learning complexity is defined as the average predicted confidence of subnets with different capacities, which encapsulates data processing within a converged model. Then we preserve the diverse and easy samples for fine-tuning. Extensive experiments with vision datasets demonstrate the effectiveness and efficiency of the proposed scoring function for classification tasks. For the instruction fine-tuning of large language models, our method achieves state-of-the-art performance with stable convergence, outperforming the full training with only 10\% of the instruction dataset.

Mitigating Privacy Risk in Membership Inference by Convex-Concave Loss

Machine learning models are susceptible to membership inference attacks (MIAs), which aim to infer whether a sample is in the training set. Existing work utilizes gradient ascent to enlarge the loss variance of training data, alleviating the privacy risk. However, optimizing toward a reverse direction may cause the model parameters to oscillate near local minima, leading to instability and suboptimal performance. In this work, we propose a novel method -- Convex-Concave Loss, which enables a high variance of training loss distribution by gradient descent. Our method is motivated by the theoretical analysis that convex losses tend to decrease the loss variance during training. Thus, our key idea behind CCL is to reduce the convexity of loss functions with a concave term. Trained with CCL, neural networks produce losses with high variance for training data, reinforcing the defense against MIAs. Extensive experiments demonstrate the superiority of CCL, achieving state-of-the-art balance in the privacy-utility trade-off.