Bo Li
Abbasi Associate Professor
Studies trustworthy machine learning, including adversarial robustness, privacy and safety of AI systems.
Bo Li's lab at UIUC works on trustworthy & safe ML, interpretability, evaluation & safety of LMs and data management & big data. 102 papers since 2023, 61 at top venues, mostly ICML and ICLR. 21 alumni. Applications close .
At a glance
- papers, 2023–2026
- 27/ year
- at T1 venues
- 60%
- PhD students
- 0
- MS students
- 0
- alumni listed
- 21
Papers by yearfrom the lab homepage, 2023–2026
- 10 papers2026 · T1 10 · collecting
- 27 papers2025 · T1 21
- 38 papers2024 · T1 27
- 27 papers2023 · T1 3
T1 venuesOtherVenues: ICML, ICLR, NeurIPS, AAAI, ACL
Research focuspaper titles that use each topic's words
- Interpretability, evaluation & safety of LMs16 papers
- Data management & big data8 papers
- Trustworthy & safe ML4 papers
- adversarial robustness1 paper
- studies trustworthy0 papers
Paper list
From the lab homepage.
202610 papers
- ICLR (T1)Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth
- ICLR (T1)ARMs: Adaptive Red-Teaming Agent against Multimodal Models with Plug-and-Play Attacks
- ICML (T1)BlueCodeAgent: A Blue Teaming Agent Powered by Automated Red Teaming for CodeGen AI
- ICML (T1)Just Ask: Curious Code Agents Reveal System Prompts in Frontier LLMs
- ICML (T1)Reasoning over Boundaries: Enhancing Specification Alignment via Test-time Deliberation
- ICLR (T1)RedCodeAgent: Automatic Red-teaming Agent against Diverse Code Agents
- ICLR (T1)Scaling Agent Learning via Experience Synthesis
- ICLR (T1)SOSBENCH: Benchmarking Safety Alignment on Scientific Knowledge
- ICLR (T1)TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models
- ICML (T1)Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents
202527 papers
- ICML (T1)AdvAgent: Controllable Blackbox Red-teaming on Web Agents
- NeurIPS (T1)Adversarial Attacks against Closed-Source MLLMs via Feature Optimal Alignment
- ICLR (T1)AdvWave: Stealthy Adversarial Jailbreak Attack against Large Audio-Language Models
- —AIR-BENCH: A Safety Benchmark based on Regulation and Policies Specified Risk Categories
- —AutoDAN-Turbo: A Lifelong Agent for Strategy Self-Exploration to Jailbreak LLMs
- NeurIPS (T1)AutoRedTeamer: Autonomous Red Teaming with Lifelong Attack Integration
- ICLR (T1)Benchmarking Vision Language Model Unlearning via Fictitious Facial Identity Dataset
- NeurIPS (T1)C-SafeGen: Certified Safe LLM Generation with Claim-Based Streaming Guardrails
- ICLR (T1)Can Watermarks be Used to Detect LLM IP Infringement For Free?
- AAAI (T1)COMMIT: Certifying Robustness of Multi-Sensor Fusion Systems against Semantic Attacks
- AAAI (T1)DiffScene: Diffusion-Based Safety-Critical Scenario Generation for Autonomous Vehicles
- —EIA: Environmental Injection Attack on Generalist Web Agents for Privacy Leakage
202438 papers
- USENIX Security (T1)ACE: A Model Poisoning Attack on Contribution Evaluation Methods in Federated Learning
- NeurIPS (T1)AgentPoison: Red-teaming LLM Agents via Memory or Knowledge Base Backdoor Poisoning
- ACL (T1)ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs
- ICLR (T1)BadChain: Backdoor Chain-of-Thought Prompting for Large Language Models
- NeurIPS (T1)BECAUSE: Bilinear Causal Representation for Generalizable Offline Model-based Reinforcement Learning
- EMNLP (T1)BEEAR: Embedding-based Adversarial Removal of Safety Backdoors in Instruction-tuned Language Models
- —C-RAG: Certified Generation Risks for Retrieval-Augmented Language Models
- ACL (T1)CaMML: Context-Aware Multimodal Learner for Large Models
- —Can Pruning Improve Certified Robustness of Neural Networks?
- CVPR (T1)ChatScene: Knowledge-Enabled Safety-Critical Scenario Generation for Autonomous Vehicles
- ICLR (T1)COLEP: Certifiably Robust Learning-Reasoning Conformal Prediction via Probabilistic Circuits
- NeurIPS (T1)Consistency Purification: Effective and Efficient Diffusion Purification towards Certified Robustness
202327 papers
- —Can Brain Signals Reveal Inner Alignment with Human Languages?
- SaTMLCARE: Certifiably Robust Learning with Reasoning via Variational Inference
- —CBD: A Certified Backdoor Detector Based on Local Dominant Probability
- —DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models
- ICLR (T1)DensePure: Understanding Diffusion Models towards Adversarial Robustness
- —DiffAttack: Evasion Attacks Against Diffusion-Based Adversarial Purification
- —DiffSmooth: Certifiably Robust Learning via Diffusion Models and Local Smoothing
- —Domain Watermark: Effective and Harmless Dataset Copyright Protection is Closed at Hand
- SaTMLEDoG: Adversarial Edge Detection For Graph Neural Networks
- SaTMLFaShapley: Fast and Approximated Shapley Based Model Pruning Towards Certifiably Robust DNNs
- —FedGame: A Game-Theoretic Defense against Backdoor Attacks in Federated Learning
- AISTATS (T1)Group Distributionally Robust Reinforcement Learning with Hierarchical Latent Variables
Where alumni went21 listed
Apply to UIUCFall 2027
UIUC Siebel Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- optional
- English
- TOEFL iBT / IELTS
- Fee
- USD 90
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USMichigan
Gabriel Poesia
- Reasoning & neuro-symbolic AI
- CHEPFL
Pascal Frossard
- Trustworthy & safe ML
- USCMU
Hoda Heidari
- Trustworthy & safe ML
- USUSC
Matteo Sesia
- Learning theory & optimization
- USUT Austin
Inderjit Dhillon
- Recognition, detection & segmentation
- CNSJTU
Wenrui Dai
- Learning theory & optimization
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
UIUCcloses