No article text was recorded for this version.
internal state monitoring
Version 1 of 132 papers41 conceptsfrozen 2026-10-04
Papers in this version
- Blog: LLMs Encode Harmfulness and Refusal Separately
- Blog: LARGO: Latent Adversarial Reflection through Gradient Optimization for Jailbreaking LLMs
- Blog: Language Models Can Predict Their Own Behavior
- Blog: Advancing LLM Safe Alignment with Safety Representation Ranking
- Blog: NExT-Guard: Training-Free Streaming Safeguard without Token-Level Labels
- Blog: RL-Obfuscation: Can Language Models Learn to Evade Latent-Space Monitors?
- Blog: Persona Vectors: Monitoring and Controlling Character Traits in Language Models
- Blog: Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth
- Blog: SafeCompass: Dynamic Chain-of-Thought Steering via Inference-Time Safety Signals
- Blog: C-SafeGen: Certified Safe LLM Generation with Claim-Based Streaming Guardrails
- Blog: AutoDAN: Automatic and Interpretable Adversarial Attacks on Large Language Models
- Blog: Exploiting Programmatic Behavior of LLMs: Dual-Use Through Standard Security Attacks
- Blog: Root Defense Strategies: Ensuring Safety of LLM at the Decoding Level
- Blog: State-Dependent Safety Failures in Multi-Turn Language Model Interaction
- Blog: ShieldHead: Decoding-time Safeguard for Large Language Models
- Blog: Aligning Large Language Models with Human Preferences through Representation Engineering
- Blog: DiffuGuard: How Intrinsic Safety is Lost and Found in Diffusion Large Language Models
- Blog: HiddenDetect: Detecting Jailbreak Attacks against Multimodal Large Language Models via Monitoring Hidden States
- Blog: GAVEL: Towards Rule-Based Safety through Activation Monitoring
- Blog: SafeSwitch: Steering Unsafe LLM Behavior via Internal Activation Signals
- Blog: CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations
- Blog: Mitigating Deep Reinforcement Learning Backdoors in the Neural Activation Space
- Blog: Efficient Detection of Toxic Prompts in Large Language Models
- Blog: PlugGuard: A Streaming Safeguard for Large Models via Latent Dynamics-Guided Risk Detection
- Blog: Intrinsic Self-correction for Enhanced Morality: An Analysis of Internal Mechanisms and the Superficial Hypothesis
- Blog: Maybe I Should Not Answer That, but... Do LLMs Understand The Safety of Their Inputs?
- Blog: Feint and Attack: Jailbreaking and Protecting LLMs via Attention Distribution Modeling
- Blog: Trojan-Speak: Bypassing Constitutional Classifiers with No Jailbreak Tax via Adversarial Finetuning
- Blog: RepIt: Steering Language Models with Concept-Specific Refusal Vectors
- Blog: Neural Chameleons: Language Models Can Learn to Hide Their Thoughts from Unseen Activation Monitors
- Blog: Beyond the Prompt: Leveraging Pre-Decoding States for Jailbreak Detection in dLLMs
- Blog: Improving Alignment and Robustness with Circuit Breakers
History & citation
Current · Permanent link
- This direction split from Latent Space Monitoring for LLM Safety (59% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_internal_state_monitoring_faf08d53,
title = {internal state monitoring},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/internal-state-monitoring/rdr_faf08d53f43441cdad597809d400bf61},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}This exact version (version 1, frozen text):
@misc{lacuna_internal_state_monitoring_faf08d53_v1,
title = {internal state monitoring},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/internal-state-monitoring/rdr_faf08d53f43441cdad597809d400bf61/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}