Zhan Su

Smooth Reading: Bridging the Gap of Recurrent LLM to Self-Attention LLM on Long-Context Understanding

Towards Fine-grained Audio Captioning with Multimodal Contextual Fusion

TTS-Hub: Leveraging Modular LoRAs and Arithmetic Composition for Controllable Text-to-Speech

CP Merging: Joint LoRA Merging using Canonical Polyadic Decomposition

It Takes Two: Your GRPO Is Secretly DPO

MARVEL: Modular Abstention for Reliable and Versatile Expert LLMs

Latent-to-Observable Score Correction for Probabilistic Time Series Imputation

Exploring Sparse Adapters for Scalable Merging of Parameter Efficient Experts