Attackers can force LLM guardrails into extended reasoning loops via optimized payloads, causing 13-63x token amplification and up to 148x latency in agent systems.
Bypassing Prompt Injection and Jailbreak Detection in LLM Guardrails
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
verdicts
UNVERDICTED 7representative citing papers
PsychoPass shows adversarial LLM conversations exhibit an early geometric fingerprint in representation space that persists after removing length confounds and is detectable from short prefixes.
Survey identifying technical and supply-chain barriers to GDPR data subject rights in ML, with new framing of 'models in the dark' for downstream opacity.
Develops a taxonomy of security interaction levels in AI/cloud infrastructure and demonstrates practical attacks exploiting isolation assumptions.
Introduces a migration framework and A/B/C taxonomy to convert static expert LLM workflows into self-evolution-ready harnesses.
Indirect prompt injection attacks remain effective on LLMs using web search tools, allowing data exfiltration and exposing ongoing weaknesses in current model defenses.
Proposes a fault-tolerance architecture for AI safety by analogizing unreliable AI artifacts to Byzantine nodes and applying consensus mechanisms.
citing papers explorer
-
From Shield to Target: Denial-of-Service Attacks on LLM-Based Agent Guardrails
Attackers can force LLM guardrails into extended reasoning loops via optimized payloads, causing 13-63x token amplification and up to 148x latency in agent systems.
-
PsychoPass: Geometric Profiling of Multi-Turn Adversarial LLM Conversations
PsychoPass shows adversarial LLM conversations exhibit an early geometric fingerprint in representation space that persists after removing length confounds and is detectable from short prefixes.
-
Short paper: Models in the dark -- Rectification and erasure under GDPR in ML supply chains
Survey identifying technical and supply-chain barriers to GDPR data subject rights in ML, with new framing of 'models in the dark' for downstream opacity.
-
Investigating The Security of Modern AI and Cloud Infrastructure
Develops a taxonomy of security interaction levels in AI/cloud infrastructure and demonstrates practical attacks exploiting isolation assumptions.
-
Toward Self-Evolution-Ready Workflow Harnesses: A Reversible Migration Path and Convertibility Taxonomy for Expert LLM Pipelines
Introduces a migration framework and A/B/C taxonomy to convert static expert LLM workflows into self-evolution-ready harnesses.
-
Exploiting Web Search Tools of AI Agents for Data Exfiltration
Indirect prompt injection attacks remain effective on LLMs using web search tools, allowing data exfiltration and exposing ongoing weaknesses in current model defenses.
-
A Byzantine Fault Tolerance Approach towards AI Safety
Proposes a fault-tolerance architecture for AI safety by analogizing unreliable AI artifacts to Byzantine nodes and applying consensus mechanisms.