Hi, there! I’m Manling. I am an assistant professor of Computer Science at Northwestern University, affiliated with Center for Robotics and Biosystems and Cognitive Science. I direct the Machine Learning and Language (MLL) Lab. I am also an Amazon Scholar working on conversational agents.
Prior to this, I was a postdoc at Stanford University (2024), working with Jiajun Wu and mentored by Fei-Fei Li.
I obtained my PhD at UIUC (2023), advised by Heng Ji and mentored by Shih-Fu Chang.
I am a recipient of MIT Tech Review 35 Under 35 in 2025, ACL Inaugural Best Desseratation Award Hornorable Mention in 2025, DARPA Riser in 2022, and a EECS Rising Star in 2022, Microsoft Research PhD Fellowship in 2021, etc. Our work on multimodal reasoning was recognized as ACL'24 Outstanding Paper Award, ACL'20 Best Demo Paper Award, NAACL'21 Best Demo Paper Award, etc. I served as organizing committee of ACL'25 (virtual co-chairs), NAACL'25 (publication co-chairs), EMNLP'24 (demo co-chairs), etc.
I aim to make AI and LLMs beneficial for humans, working at the intersection of language, vision, robotics, and society. I focus on Reasoning, Planning, Compositionality, with applications in Embodied AI and AI for science. At the core of my research, I build machines that interact with the physical world via multimodal data (Language + X, where X can be robotics, vision, audio, etc). The ultimate goal is to promote trustworthiness and truthfulness through a structured view that is explainable, highly compositional, and capable of long-horizon reasoning.
-
Agent Model Training, Planning and Reasoning.
- Foundation Agent Training:
LLM Agent (RAGEN, Self-Play Agent),
VLM Agent (VAGEN with World Model RL)
- Embodied Agents:
Embodied Reasoning Agent for RL training,
Embodied Agent Interface for LLM agents,
EmbodiedBench for VLM agents,
ROSETTA with human feedback,
IKEA Manual at Work for 4D grouded planning.
- Planning, Reasoning and Compositionality:
Planning Logic,
Procedural Planning,
Agriculture Task Planning,
LLM Schema,
Graph Schema,
Generative Graph Schema,
PathLM,
Decomposition to Latent Text Prompt.
-
Spatial Intelligence, Multimodal (Language + Vision/Robotics).
- Spatial Reasoning:
MindCube,
AdaptVis,
Vector Graphics Reasoning,
LayoutVLM.
- World Modeling:
VAGEN within RL,
Self-Play Agent before RL,
Embodied Agent Interface with BDDL transition models
- Long-Horizon multimodal intelligence:
T* for temporal search,
HourVideo,
LM4Video,
VideoArgument,
VideoEvent,
- Multimodal alignment:
CLIP-Event,
VisualDecompsition,
Knowledge-Driven Vision-Language Pretraining,
VisualKnowledge,
MuMuQA,
M2E2,
Image2Code.
-
Mechanistic Interpretability.
- Open up models to understand inner workings:
Why is Spatial Reasoning Hard for VLMs?,
Reasoning/Perception Merging,
Exploring Diffusion Transformer Designs via Grafting.
- Physics of LLMs/VLMs:
Knowledge Overshadow,
Ripple Effect,
- Control and intervene foundation models:
LM-Steer,
Hallu-Control,
Deep Concept Injection.
-
Factualness, Trustworthiness.
- NLP + Human/Social (theory of mind, debiasing, propoganda, knowledge graph, information extraction):
SmartBook,
Explainable Fact Checking,
Timeline Summarization,
Meeting Summarization,
GAIA IE,
Info Propagation Pattern,
DebiasPrompt,
Multilingual KG,
Object Hallucination.
- AI for Science, claim feasibility verification, figure/chart editing, paper/review generation:
COVID-19 Claim Radar,
COVID Knowledge Graph.
Prospective students: I have several PhD positions in Fall 2026 and intern positions.
» Show details
>> We have
amazing new faculty at NU working on robotics and foundation models, please check out
Ruohan! He also has several PhD positions in Fall 2026.
>> If you are interested in PhD at Northwestern, please apply to
NorthwesternCS. Due to the large amount of emails,
I aplogize that I will not able to reply to individual emails (Please note that
non-reply does NOT indicate non-interesting, largely means emails got missed or I unfortunately did not get time to check such emails before applications). Please
choose me as a potential advisor in the application, and I will
check every application carefully in late Dec and do interviews in Jan-Mar.
>> If you are interested in doing research internship with our group, please feel free to talk to any of the PhD students to join their projects, and many of them are looking for collaborators. The best way is to submit this
form and drop an email to
limanling.ai@gmail.com. This mailbox has been checked more frequently.
>> Prospective_Students_English
>> Prospective_Students_Chinese
>> Proud of what our students have achieved in the very first year!
Junior PhD/master/undergraduate students: I will dedicate 30 minutes each week to offer guidance/suggestions/mentorship, especially for students from underrepresented groups or whoever is in need. If you would like to chat about life, career path, graduate school applications, or research ideas related to AI/ML, feel free to file the
form to schedule a meeting.