Yitao Liang
Assistant Professor
Researches machine learning, neuro-symbolic integration and reinforcement learning with interpretable model representations.
Yitao Liang's lab at Peking works on reasoning & neuro-symbolic AI, reinforcement learning and interpretability, evaluation & safety of LMs. 46 papers since 2023, 32 at top venues, mostly ICLR and NeurIPS. Applications close .
At a glance
- papers, 2023–2026
- 12/ year
- at T1 venues
- 70%
- PhD students
- —
- MS students
- —
- alumni listed
- —
Papers by yearfrom DBLP, 2023–2026
- 12 papers2026 · T1 8 · collecting
- 19 papers2025 · T1 11
- 9 papers2024 · T1 8
- 6 papers2023 · T1 5
T1 venuesOtherVenues: ICLR, NeurIPS, ICML, AAAI, CVPR
Research focuspaper titles that use each topic's words
- Reinforcement learning6 papers
- Reasoning & neuro-symbolic AI4 papers
- Interpretability, evaluation & safety of LMs1 paper
- neuro-symbolic0 papers
- integration0 papers
Paper list
From DBLP.
202612 papers
- ISSCCA 28nm Mode-Reconfigurable CAM-CIM Hybrid Complete 3-SAT Solver Supporting Conflict-Driven Clause Learning with 100% Solvability (opens in a new tab)
- Turkish J. Electr. Eng. Comput. Sci.A deep learning-based real-time no-reference image decolorization network with perceptual preservation (opens in a new tab)
- Signal Process. Image Commun.A two-stage low-light image enhancement network combining spatial and frequency information (opens in a new tab)
- ICLR (T1)Can Language Models Discover Scaling Laws? (opens in a new tab)
- ASPLOS (T1)CREATE: Cross-Layer Resilience Characterization and Optimization for Efficient yet Reliable Embodied AI Systems (opens in a new tab)
- AAAI (T1)DexGraspVLA: A Vision-Language-Action Framework Towards General Dexterous Grasping (opens in a new tab)
- HPCA (T1)ESTroM: Element-Flow Architecture for Processing Sparse Tractable Probabilistic Models (opens in a new tab)
- ICLR (T1)Inference-time scaling of diffusion models through classical search (opens in a new tab)
- TPAMI (T1)On the Limitations and Capabilities of Position Embeddings for Length Generalization (opens in a new tab)
- IEEE J. Solid State CircuitsSKADI: A 28-nm Complete K-SAT Solver Featuring Bidirectional In-Memory Deduction and Incremental Updating (opens in a new tab)
- AAAI (T1)Steering Visuomotor Policy in Open Worlds via Cross-View Goal Alignment (opens in a new tab)
- CVPR (T1)Training One Model to Master Cross-Level Agentic Actions via Reinforcement Learning (opens in a new tab)
202519 papers
- ISSCC37.5 SKADI: A 28nm Complete K-SAT Solver Featuring Dual-Path SRAM-Based Macro and Incremental Update with 100% Solvability (opens in a new tab)
- J. Supercomput.CELAT-DiffNet: channel-enhanced local attention transformer for underwater image enhancement based on diffusion models (opens in a new tab)
- ICLR (T1)GROOT-2: Weakly Supervised Multimodal Instruction Following Agents (opens in a new tab)
- Earth Sci. InformaticsHiLoF-GAN: high and low frequency information separation GAN for underwater image enhancement (opens in a new tab)
- TPAMI (T1)JARVIS-1: Open-World Multi-Task Agents With Memory-Augmented Multimodal Language Models (opens in a new tab)
- Findings of ACLJARVIS-VLA: Post-Training Large-Scale Vision Language Models to Play Visual Games with Keyboards and Mouse (opens in a new tab)
- ICML (T1)Low-Dimension-to-High-Dimension Generalization and Its Implications for Length Generalization (opens in a new tab)
- ICML (T1)MCU: An Evaluation Framework for Open-Ended Game Agents (opens in a new tab)
- Circuits Syst. Signal Process.MDFF-Net: Multi-level Dynamic Feature Fusion Network Combined with Deep Supervision Mechanism for Low-Light Image Enhancement (opens in a new tab)
- ICCV (T1)Open-World Skill Discovery from Unsegmented Demonstration Videos (opens in a new tab)
- NeurIPS (T1)Plug-and-Play Context Feature Reuse for Efficient Masked Generation (opens in a new tab)
- NeurIPS (T1)QCircuitBench: A Large-Scale Dataset for Benchmarking Quantum Algorithm Design (opens in a new tab)
20249 papers
- NeurIPS (T1)A Tractable Inference Perspective of Offline RL (opens in a new tab)
- ICLR (T1)GROOT: Learning to Follow Instructions by Watching Gameplay Videos (opens in a new tab)
- NeurIPS (T1)Mars: Situated Inductive Reasoning in an Open-World Environment (opens in a new tab)
- ICLR (T1)Neural-Symbolic Recursive Machine for Systematic Generalization (opens in a new tab)
- NeurIPS (T1)OmniJARVIS: Unified Vision-Language-Action Tokenization Enables Open-World Instruction Following Agents (opens in a new tab)
- AAAI (T1)ProAgent: Building Proactive Cooperative Agents with Large Language Models (opens in a new tab)
- Findings of ACLRulE: Knowledge Graph Reasoning with Rule Embedding (opens in a new tab)
- ICML (T1)Selecting Large Language Model to Fine-tune via Rectified Scaling Law (opens in a new tab)
- NeurIPS (T1)TFG: Unified Training-Free Guidance for Diffusion Models (opens in a new tab)
20236 papers
- NeurIPS (T1)Describe, Explain, Plan and Select: Interactive Planning with LLMs Enables Open-World Multi-Task Agents (opens in a new tab)
- Mach. Vis. Appl.Effective long-term tracking with contrast optimizer (opens in a new tab)
- CVPR (T1)Open-World Multi-Task Control Through Goal-Aware Representation Learning and Adaptive Horizon Prediction (opens in a new tab)
- AAAI (T1)Out-of-Distribution Generalization by Neural-Symbolic Joint Training (opens in a new tab)
- ICLR (T1)SQA3D: Situated Question Answering in 3D Scenes (opens in a new tab)
- ICML (T1)Understanding the Distillation Process from Deep Generative Models to Tractable Probabilistic Circuits (opens in a new tab)
Where alumni went
Alumni are not listed on the lab homepage.
Apply to PekingFall 2027
Computer Science PhD (international applicants)
- Round
- Main
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNZhejiang
Yubo Tao
- Visualization & visual analytics
- CNPeking
Yonghong Tian
- Computational neuroscience & cognition
- USUSC
Evi Micha
- Trustworthy & safe ML
- CNZhejiang
Yin Zhang
- Generative & foundation models
- USMIT
Sendhil Mullainathan
- Human-AI collaboration & trust
- USUCSD
Berk Ustun
- Learning theory & optimization
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The member list could not be found on the lab site.
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- DBLP
- Members
- Not collected
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
Pekingcloses