arXiv AI
Aug 5
Your Agentic LLMs Secretly Encode Latent Signals of Indirect Prompt-Injection Exposure
arXiv:2608. 02657v1 Announce Type: cross Abstract: Agentic LLMs are vulnerable to indirect prompt injection (IPI) attacks, e.
Visible Reasoning and Indirect Prompt-Injection Monitorability Across English, Tamil, and Tanglish
arXiv:2608. 15392v1 Announce Type: new Abstract: Chain-of-thought monitoring is a potentially useful safety signal, but its reliability across languages and behavioral settings remains uncertain.
Related stories
arXiv AI
Jun 10
When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models
arXiv:2606. 10740v1 Announce Type: new Abstract: Failures in multi-turn reasoning models are largely invisible to terminal-score evaluation.
arXiv AI
Jun 19
Analyzing the Narration Gap in LLM-Solver Loops
arXiv:2606. 19588v1 Announce Type: new Abstract: Formal tools such as SAT and SMT solvers are increasingly embedded in language model reasoning pipelines when a safety or security critical question can be formulated in logic.
arXiv AI
Aug 5
The Tell-Tale Trace: Detecting Reasoning Failures in LLMs Using Chain-of-Thought Dynamics
arXiv:2608. 03291v1 Announce Type: cross Abstract: Chain-of-thought (CoT) reasoning improves large language model (LLM) performance while also providing an observable interface to the model's reasoning process.
arXiv AI
Aug 5
Evading Chain-of-Thought Monitoring Through Model Poisoning
arXiv:2608. 02820v1 Announce Type: cross Abstract: Chain-of-thought (CoT) monitoring is an increasingly important component of AI safety stacks but relies on the assumption that a model's reasoning trace is informative about its actions.
arXiv Machine Learning
Aug 11
BASIS: Breach-Aware Selective Prompt Injection Shielding with Prefill Attention Probes
arXiv:2608. 08027v1 Announce Type: cross Abstract: Prompt injection is a critical security threat in large language model (LLM) applications, where attackers hijack model behavior by embedding malicious instructions in user or external data.
arXiv AI
Jul 21
Reasoning as a Double-Edged Sword: Architecture and Cross-Stage Robustness in Vision-Language-Action Models
arXiv:2607. 17786v1 Announce Type: cross Abstract: Does adding a reasoning step make a Vision-Language-Action (VLA) model more robust to perturbation?
arXiv AI
Jun 9
RecurGuard: Runtime Monitoring for Reasoning-Token Consumption Attacks
arXiv:2606. 07968v1 Announce Type: cross Abstract: Reasoning-capable large language models can be induced to spend their generation budget on injected decoy tasks rather than answering the user's question, causing denial of service when no final answer is produced and denial of wallet when excess output tokens are billed.
arXiv Machine Learning
Jun 18
Evaluating Prompting-Based Defenses Against Domain-Camouflaged Injection Attacks
arXiv:2606. 18530v1 Announce Type: cross Abstract: Domain-camouflaged injection attacks embed malicious instructions in retrieved content using domain-appropriate vocabulary, evading standard detectors that rely on syntactic injection markers.
arXiv AI
Jun 26
HauntAttack: When Attack Follows Reasoning as a Shadow
arXiv:2506. 07031v5 Announce Type: replace-cross Abstract: Emerging Large Reasoning Models (LRMs) consistently excel in mathematical and reasoning tasks, showcasing remarkable capabilities.
arXiv AI
Jul 22
CPInj: Uncovering Prompt Injection Risks in Textual Collaborative Prompt Optimization
arXiv:2607. 18622v1 Announce Type: cross Abstract: Textual Collaborative Prompt Optimization (TCPO) extends Textgrad (Yuksekgonul et al.
arXiv Machine Learning
Jun 25
Do Thinking Tokens Help with Safety?
arXiv:2606. 25013v1 Announce Type: new Abstract: Today's reasoning models use thinking tokens to attain stronger performance on benchmarks than their instruction-tuned counterparts.