HARMONY: Enhancing Hybrid Split Federated Learning Accuracy
Discover how HARMONY bridges personalization and generalization gaps in federated learning by mitigating representation skew in heterogeneous client models...
Effective Hallucination Detection Using Proxy Analyzers
Discover how proxy-analyzer activations improve hallucination detection in large language models, boosting accuracy across multiple architectures.
Fine-Tuning LLMs with Synthetic Data for Gaming Toxicity
Discover how synthetic data boosts large language models for multi-class toxicity detection in gaming chat, improving moderation and online safety.
Closed-Form Linear-Probe Dataset Distillation for Vision Models
Discover CLP-DD, a fast, efficient dataset distillation method for pre-trained vision models, boosting performance with less GPU memory and compute.
Text Uncanny Valley: LLM Performance Drop on Corrupted Text
Explore how Large Language Models face non-linear performance drops with corrupted text, revealing the Text Uncanny Valley effect in information retrieval.