David Wagner
Professor
Studies computer security, including the security of machine learning systems and usable security.
David Wagner's lab at UC Berkeley works on trustworthy & safe ML. 22 papers since 2023, 10 at top venues, mostly ICLR and USENIX Security. 45 alumni. Applications close .
At a glance
- papers, 2023–2026
- 6/ year
- at T1 venues
- 45%
- PhD students
- —
- MS students
- —
- alumni listed
- 45
Papers by yearfrom the lab homepage, 2023–2026
- 3 papers2026 · T1 1 · collecting
- 10 papers2025 · T1 3
- 3 papers2024 · T1 2
- 6 papers2023 · T1 4
T1 venuesOtherVenues: ICLR, USENIX Security, CCS, ICCV, NAACL
Research focuspaper titles that use each topic's words
- Trustworthy & safe ML1 paper
- security0 papers
- computer security0 papers
- learning systems0 papers
- studies computer0 papers
Paper list
From the lab homepage.
20263 papers
202510 papers
- AISecDefending Against Prompt Injection With a Few DefensiveTokens (opens in a new tab)
- AISecJailbreaksOverTime: Detecting Jailbreak Attacks Under Distribution Shift (opens in a new tab)
- SaTMLMark My Words: Analyzing and Evaluating Language Model Watermarks (opens in a new tab)
- SaTMLML-Based Behavioral Malware Detection Is Far From a Solved Problem (opens in a new tab)
- ACM CODASPYPromptShield: Deployable Detection for Prompt Injection Attacks (opens in a new tab)
- CCS (T1)SecAlign: Defending Against Prompt Injection with Preference Optimization (opens in a new tab)
- LLM4SecSeedAIchemy: LLM-driven Seed Corpus Generation for Fuzzing (opens in a new tab)
- NAACL (T1)Stronger Universal and Transferable Attacks by Suppressing Refusals (opens in a new tab)
- USENIX Security (T1)StruQ: Defending Against Prompt Injection with Structured Queries (opens in a new tab)
- International Conference on Software Engineering (ICSE)Vulnerability Detection with Code Language Models: How Far Are We? (opens in a new tab)
20243 papers
20236 papers
- USENIX Security (T1)Continuous Learning for Android Malware Detection (opens in a new tab)
- Second Workshop on Situating Network Infrastructure with People, Practices, and BeyondDeTagTive: Linking MACs to Protect Against Malicious BLE Trackers (opens in a new tab)
- RAIDDiverseVul: A New Vulnerable Source Code Dataset for Deep Learning Based Vulnerability Detection (opens in a new tab)
- USENIX Security (T1)Network Detection of Interactive SSH Impostors Using Deep Learning (opens in a new tab)
- ICLR (T1)Part-Based Models Improve Adversarial Robustness (opens in a new tab)
- ICCV (T1)REAP: A Large-Scale Realistic Adversarial Patch Benchmark (opens in a new tab)
Where alumni went45 listed
Apply to UC BerkeleyFall 2027
UC Berkeley EECS Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- not required
- English
- TOEFL/IELTS (non-English-country degrees)
- Fee
- See program page
Confirmed on the program page.
Similar labsnearest by research summary
- CNZhejiang
Feng Lin
- Trustworthy & safe ML
- USUC Berkeley
Anil Aswani
- Learning theory & optimization
- CNSJTU
Yunbo Wang
- Reinforcement learning
- USCMU
Aarti Singh
- Reinforcement learning
- HKCUHK
Xiangyu Yue
- Vision-language & multimodal
- USMichigan
Yixin Wang
- Probabilistic & causal inference
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 lab site and the paper record do not agree.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Not collected
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
UC Berkeleycloses