transferability
Transferability describes the ability of a trained model to perform well on new tasks or domains that it was not explicitly trained on. This concept is important for creating robust AI systems that can generalize knowledge and skills acquired from one context to another.
- A Frustratingly Simple Yet Highly Effective Attack Baseline: Over 90% Success Rate Against the Strong Black-box Models of GPT-4.5/4o/o1
- BAM-ICL: Causal Hijacking In-Context Learning with Budgeted Adversarial Manipulation
- Boosting Adversarial Transferability with Spatial Adversarial Alignment
- Consensus-Robust Transfer Attacks via Parameter and Representation Perturbations
- Dual-Flow: Transferable Multi-Target, Instance-Agnostic Attacks via $\textit{In-the-wild}$ Cascading Flow Optimization
- FEEL: Quantifying Heterogeneity in Physiological Signals for Generalizable Emotion Recognition
- Fit the Distribution: Cross-Image/Prompt Adversarial Attacks on Multimodal Large Language Models
- Flatten Graphs as Sequences: Transformers are Scalable Graph Generators
- GC4NC: A Benchmark Framework for Graph Condensation on Node Classification with New Insights
- Harnessing the Computation Redundancy in ViTs to Boost Adversarial Transferability
- JAMUN: Bridging Smoothed Molecular Dynamics and Score-Based Learning for Conformational Ensemble Generation
- Latency NMS Attacks: Is It Real Life or Is It Just Fantasy?
- MLIP Arena: Advancing Fairness and Transparency in Machine Learning Interatomic Potentials via an Open, Accessible Benchmark Platform
- Meta-learning how to Share Credit among Macro-Actions
- MiNT: Multi-Network Transfer Benchmark for Temporal Graph Learning
- Non-Adaptive Adversarial Face Generation
- On Transferring Transferability: Towards a Theory for Size Generalization
- SAINT: Sequence-Aware Integration for Spatial Transcriptomics Multi-View Clustering
- Security Challenges in AI Agent Deployment: Insights from a Large Scale Public Competition
- System Prompt Optimization with Meta-Learning
- TransferBench: Benchmarking Ensemble-based Black-box Transfer Attacks
- TransferTraj: A Vehicle Trajectory Learning Model for Region and Task Transferability
- TransferTraj: A Vehicle Trajectory Learning Model for Region and Task Transferability
- Transformer Copilot: Learning from The Mistake Log in LLM Fine-tuning
- Versatile Transferable Unlearnable Example Generator