arXiv AI
Jun 9
Prescriptive Scaling Reveals the Evolution of Language Model Capabilities
arXiv:2602. 15327v2 Announce Type: replace-cross Abstract: Machine learning model performance improvements tend to arise from competition and application.
Caliber: Cross-Architecture Extraction-Cost Control for Score-Returning APIs
arXiv:2608. 01023v1 Announce Type: new Abstract: We present Caliber, an output-perturbation defense against model extraction that formulates noise selection as a calibration problem: how much the defense degrades the supervision signal used to train a surrogate, and the provable per-input query cost of recovering the clean logits.
Related stories
arXiv Machine Learning
Jun 3
RRISE: Robust Radius Inference via a Surrogate Estimator
arXiv:2606. 02876v1 Announce Type: new Abstract: Randomized smoothing (RS) uses a smoothed classifier to provide architecture-agnostic certificates of $\ell_2$ classification robustness, but its dependence on per-input Monte Carlo (MC) sampling undermines its use in real-time systems.
arXiv Machine Learning
Jun 17
MorphStrata: Layer-Specific Perturbations for Generating Morphence Students in Time-Series Moving Target Defense
arXiv:2606. 17435v1 Announce Type: new Abstract: Time-series forecasting models remain vulnerable to gradient-based adversarial attacks while existing defense mechanisms typically incur a trade-off in robustness for bounded response and compute cost.
arXiv AI
Jun 10
Is Fairness Truly Fair? Towards Reliable Lipschitz Fairness in Multi-Task Learning via Fixed-\texorpdfstring{$\delta$}{delta} Alignment
arXiv:2606. 10632v1 Announce Type: cross Abstract: Lipschitz-style individual fairness formalizes the idea that semantically similar examples should receive similar predictions, but its evaluation in multi-task learning (MTL) can be confounded by method-induced representation scales.
arXiv Machine Learning
Jun 16
Model Stealing Through the Lens of Model Multiplicity
arXiv:2606. 15493v1 Announce Type: new Abstract: Model stealing attacks, where adversaries create high-fidelity surrogate models, are a significant threat to the intellectual property of machine learning services.
arXiv AI
Jul 7
Incompressible Knowledge Probes: Estimating Black-Box LLM Parameter Counts via Factual Capacity
arXiv:2604. 24827v2 Announce Type: replace-cross Abstract: Closed-source frontier labs do not disclose parameter counts.
arXiv AI
Jul 28
UNIFUSION: Adapting Autoregressive Language Models into Discrete Diffusion under a Unified Reverse-Rate Objective
arXiv:2607. 24507v1 Announce Type: cross Abstract: Existing methods mainly adapt pretrained autoregressive (AR) language models to masked diffusion, whereas we directly adapt them to uniform-noise diffusion, where every token remains editable during sampling.
arXiv Machine Learning
Jun 17
AIMER: Calibration-Free Task-Agnostic MoE Expert Pruning
arXiv:2603. 18492v3 Announce Type: replace Abstract: Mixture-of-Experts (MoE) language models increase parameter capacity without proportional per-token computation, yet deployment still requires storing the full expert pool, making expert pruning important for reducing memory and serving overhead.
arXiv AI
1d ago
Measuring Reward Hacking and Reasoning-Answer Decoupling Under Position-Confounded Optimization
arXiv:2608. 15445v1 Announce Type: new Abstract: When a reward is correct on every training example yet consistent with more than one goal, a model can acquire an unintended one, a failure known as goal misgeneralization.
Hugging Face Trending Papers
Jul 23
Mean-to-Score Discrete Diffusion: Posterior-Mean Denoisers for Score Entropy
Score Entropy Discrete Diffusion (SEDD) parameterizes discrete reverse processes with unconstrained positive score ratios. While positivity guarantees nonnegative reverse jump rates, it does not ensure Bayes realizability: ratios at a noisy state need not be jointly induced by any clean-token posterior under the forward kernel.
arXiv AI
Jun 3
q0: Primitives for Hyper-Epoch Pretraining
arXiv:2606. 03938v1 Announce Type: cross Abstract: Multi-epoch training is becoming the standard now that compute is growing faster than the supply of high-quality text.
arXiv AI
Jul 3
LACUNA: A Testbed for Evaluating Localization Precision for LLM Unlearning
arXiv:2607. 02513v1 Announce Type: cross Abstract: LLMs memorize sensitive training data, including personally identifiable information (PII), creating a pressing need for reliable post hoc removal methods.