← Search

Zhengping Ji

2 accepted papers

2026

LiME: Lightweight Mixture of Experts for Efficient Multimodal Multi-task Learning

ICML 2026spotlight

MoE-PEFT methods combine Mixture of Experts with parameter-efficient fine-tuning for multi-task adaptation, but require separate adapters per expert—causing trainable parameters to scale linearly with expert count and limiting applicability to adapter-based architectures. We propose LiME (Lightweigh…

Cited by 0SourceScholar
2025

DHP Benchmark: Are LLMs Good NLG Evaluators?

NAACL 2025findings

Large Language Models (LLMs) are increasingly serving as evaluators in Natural Language Generation (NLG) tasks; this is often referred to as “LLM-as-a-judge” paradigm. However, the capabilities of LLMs in evaluating NLG quality remain underexplored. Current studies depend on human assessments and si…

Cited by 6SourcePDFScholar