arXiv Machine Learning
Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spots
Read the original on arXiv Machine Learning →The Flow has not summarised this story yet — read it at arXiv Machine Learning.
Related stories
arXiv Machine Learning
Aug 14
Gradient-Free Warm-Start Library Recovery: an Amortized-Regret Separation
arXiv:2606. 21253v2 Announce Type: replace Abstract: Continual learning that is gradient-free, local, online, and append-only is attractive for edge and streaming deployment, but its value is usually argued informally.
arXiv Statistics ML
3d ago
An Exponential Deterministic--Randomized Gap in ERM-Oracle Complexity for Thresholds on an Unknown Order
arXiv:2609.10196v1 Announce Type: cross Abstract: Attias, Hanneke and Ramaswami (NeurIPS 2025) asked whether randomization provably reduces the oracle calls needed for online learning when the class...
arXiv Machine Learning
Jun 2
Auditing Near-Optimal Policies Can Be Exponentially Hard: Conditional Query Lower Bounds via Occupancy Rashomon Capacity
arXiv:2606. 00414v1 Announce Type: new Abstract: When many reinforcement-learning policies achieve near-optimal return, a post-hoc auditor may have to distinguish among many behaviorally distinct but return-equivalent policies.
arXiv Machine Learning
Aug 7
Optimal Rates for Learning with Monotone Adversaries
arXiv:2608. 06337v1 Announce Type: cross Abstract: A monotone adversary observes an i.
arXiv Machine Learning
Aug 14
Thermodynamics of Learning: A Typed Four-Component Accounting of Memory, Fit, and Value
arXiv:2608. 12791v1 Announce Type: cross Abstract: What a finite learning device has recorded and what will hold value for it on future tasks are not the same quantity.
arXiv AI
3d ago
PAC-Private Autoregressive Generation: Calibrating Noise to Ensemble Disagreement
The paper introduces PAC‑Private Autoregressive Generation, a method that calibrates noise based on ensemble disagreement across overlapping ‘worlds’ of a private corpus, thereby extending PAC privacy from classification to text generation. By training adapters on a frozen public model and using posterior‑weighted disagreement to add noise only when predictions vary, the approach achieves strong privacy guarantees while preserving most of the fine‑tuning benefit. Experiments on WikiText‑103 with GPT‑2‑small show 74 % of the fine‑tuning gain retained with a per‑token budget of 2⁻³², and membership‑inference success bounded to 51.08 % after one million tokens, outperforming PMixED under matched conditions.