| From Trojan Horses to Castle Walls: Unveiling Bilateral Data Poisoning Effects in Diffusion Models | Nov 4, 2023 | Backdoor Attackbackdoor defense | CodeCode Available | 0 |
| Bridging Lottery Ticket and Grokking: Understanding Grokking from Inner Structure of Networks | Oct 30, 2023 | Image ClassificationMemorization | CodeCode Available | 0 |
| The statistical thermodynamics of generative diffusion models: Phase transitions, symmetry breaking and critical instability | Oct 26, 2023 | MemorizationVariational Inference | —Unverified | 0 |
| Grokking in Linear Estimators -- A Solvable Model that Groks without Understanding | Oct 25, 2023 | Memorization | —Unverified | 0 |
| SoK: Memorization in General-Purpose Large Language Models | Oct 24, 2023 | MemorizationQuestion Answering | —Unverified | 0 |
| MoPe: Model Perturbation-based Privacy Attacks on Language Models | Oct 22, 2023 | Language ModellingMemorization | —Unverified | 0 |
| Implications of Annotation Artifacts in Edge Probing Test Datasets | Oct 20, 2023 | Memorization | CodeCode Available | 0 |
| Copyright Violations and Large Language Models | Oct 20, 2023 | Memorization | CodeCode Available | 0 |
| ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks | Oct 19, 2023 | HallucinationHallucination Evaluation | —Unverified | 0 |
| Training Dynamics of Deep Network Linear Regions | Oct 19, 2023 | Memorization | —Unverified | 0 |
