Ling Liu

Full Professor@Georgia Institute of Technology·美国·OpenReview

研究方向

Multi-modal learning · Large Language Models · Reinforcement Learning · Big Data Systems · Trust · Privacy · Security · Vision-Language-Action Models · Audio · Image · Text Understanding · Deep Neural Networks · Generative AI

Ling Liu

MolLangBench: A Comprehensive Benchmark for Language-Prompted Molecular Structure Recognition, Editing, and Generation

Dynamic Optimizations of LLM Ensembles with Two-Stage Reinforcement Learning Agents

A Neurosymbolic Agent System for Compositional Visual Reasoning

Adverseness vs. Equilibrium: Exploring Graph Adversarial Resilience through Dynamic Equilibrium

Booster: Tackling Harmful Fine-tuning for Large Language Models via Attenuating Harmful Perturbation

Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning Attack

PokéLLMon: A Grounding and Reasoning Benchmark for Large Language Models in Pokémon Battles