Lattice AI Research

Research focus

Red-Teaming & Adversarial Robustness (3)Tool Use & Agents (3)Open-Source Models & Weights (2)Code Generation & Program Synthesis (1)

Frequent co-authors

A. Happe (2)Benjamin Probst (1)Jürgen Cito (1)Philipp Normann (1)

Papers (3)

Apr 29, 2026

3w ago

Enhancing Linux Privilege Escalation Attack Capabilities of Local LLM Agents

Local LLMs can now rival cloud-based giants like GPT-4o in Linux privilege escalation tasks, thanks to targeted system-level and prompting interventions.

Benjamin Probst, Andreas Happe, Jürgen Cito

Open-Source Models & Weights Red-Teaming & Adversarial Robustness Tool Use & Agents

Mar 18, 2026

Philipp Normann +4Mar 18, 2026·also TU Wien

Post-Training Local LLM Agents for Linux Privilege Escalation with Verifiable Rewards

A 4B parameter model can nearly match the privilege escalation performance of a state-of-the-art closed LLM like Claude Opus, while being fully local and 100x cheaper to run.

Philipp Normann, Andreas Happe, A. Happe +2

Code Generation & Program Synthesis Open-Source Models & Weights Red-Teaming & Adversarial Robustness+1

Mar 2, 2026

Can LLMs Hack Enterprise Networks? -- Replicated Computational Results (RCR) Report

LLMs might be surprisingly capable penetration testers for enterprise networks, and this report provides the recipe to reproduce the original findings.

A. Happe, Andreas Happe, Jürgen Cito

Eval Frameworks & Benchmarks Red-Teaming & Adversarial Robustness Tool Use & Agents

Search

Andreas Happe

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (3)