black-box models
Black-box models are AI systems whose internal workings are not transparent or interpretable, making it challenging to understand how they arrive at specific decisions or outputs.
- Bayesian Concept Bottleneck Models with LLM Priors
- ConfTuner: Training Large Language Models to Express Their Confidence Verbally
- Contimask: Explaining Irregular Time Series via Perturbations in Continuous Time
- Enhancing Visual Prompting through Expanded Transformation Space and Overfitting Mitigation
- Graph-based Symbolic Regression with Invariance and Constraint Encoding
- Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMs
- PRESTO: Preimage-Informed Instruction Optimization for Prompting Black-Box LLMs
- PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding
- SHAP values via sparse Fourier representation
- Towards Building Model/Prompt-Transferable Attackers against Large Vision-Language Models
- Tree Ensemble Explainability through the Hoeffding Functional Decomposition and TreeHFD Algorithm