Skip to content

Home Search Authors Pakistan News Saved

SciRah

SciRah does not host or store research papers. Results come live from free, official scholarly APIs, and every paper links back to its original source.

Data sources

OpenAlex
Semantic Scholar
Crossref
CORE
arXiv
PubMed

Privacy policy Terms of service

Andrew Paverd · SciRah

Researcher profile · Semantic Scholar

Andrew Paverd

Share profile

Andrew Paverd — researcher profile on SciRah

h-index: 6
Works: 11
Citations: 412

Most-cited papers

Securing AI Agents with Information-Flow Control
Manuel Costa, Boris Köpf, Aashish Kolluri, Andrew Paverd +5 more
2025arXiv.org165 citations
Semantic Scholar
Design Patterns for Securing LLM Agents against Prompt Injections
Luca Beurer-Kellner, Beat Buesser, Ana-Maria Creţu, Edoardo Debenedetti +10 more
2025arXiv.org107 citations
Semantic Scholar
Get My Drift? Catching LLM Task Drift with Activation Deltas
Sahar Abdelnabi, Aideen Fay, Giovanni Cherubin, Ahmed Salem +2 more
20242025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)84 citations Open access
Semantic Scholar
LLMail-Inject: A Dataset from a Realistic Adaptive Prompt Injection Challenge
Sahar Abdelnabi, Aideen Fay, Ahmed Salem, Egor Zverev +21 more
2025arXiv.org33 citations
Semantic Scholar
Permissive Information-Flow Analysis for Large Language Models
Shoaib Ahmed Siddiqui, Radhika Gaonkar, Boris Köpf, David Krueger +6 more
2024Trans. Mach. Learn. Res.26 citations
Semantic Scholar
Closed-Form Bounds for DP-SGD against Record-level Inference Attacks
Giovanni Cherubin, Microsoft Boris Köpf, Azure Research, Andrew Paverd +5 more
9 citations
Semantic Scholar
Position: Stateless Yet Not Forgetful: Implicit Memory as a Hidden Channel in LLMs
Ahmed Salem, Andrew Paverd, Sahar Abdelnabi
20262026 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)4 citations
Semantic Scholar
Highlight & Summarize: RAG without the jailbreaks
Giovanni Cherubin, Andrew Paverd
2025arXiv.org3 citations
Semantic Scholar
MetaBackdoor: Exploiting Positional Encoding as a Backdoor Attack Surface in LLMs
Rui Wen, M. Russinovich, Andrew Paverd, J. Sakuma +1 more
2026arXiv.org2 citations
Semantic Scholar
SpecGuard: Inference-Time Backdoor Detection For Free
Wen Rui, Ahmed Salem, Andrew Paverd, M. Russinovich +1 more
20260 citations
Semantic Scholar
This paper is included in the Proceedings of the 33rd USENIX Security
Giovanni Cherubin, Microsoft Boris Köpf, Azure Research, Andrew Paverd +5 more
0 citations
Semantic Scholar

Frequent co-authors

Invite them to SciRah by sharing their profile.

Giovanni Cherubin×5
Invite
Giovanni Cherubin — researcher profile on SciRah
M. Russinovich×4
Invite
M. Russinovich — researcher profile on SciRah
Shruti Tople×4
Invite
Shruti Tople — researcher profile on SciRah
Lukas Wutschitz×4
Invite
Lukas Wutschitz — researcher profile on SciRah

×3

Invite

Ahmed Salem — researcher profile on SciRah

Santiago Zanella-Béguelin×3

Invite

Santiago Zanella-Béguelin — researcher profile on SciRah

Invite

Boris Köpf — researcher profile on SciRah

Sahar Abdelnabi×2

Invite

Sahar Abdelnabi — researcher profile on SciRah

Invite

Aideen Fay — researcher profile on SciRah

Invite

Ahmed Salem — researcher profile on SciRah

Microsoft Boris Köpf×2

Invite

Microsoft Boris Köpf — researcher profile on SciRah

Azure Research×2

Invite

Azure Research — researcher profile on SciRah