No article text was recorded for this version.
latent safety awareness
Version 1 of 122 papers26 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Unintended Misalignment from Agentic Fine-Tuning: Risks and Mitigation
- Blog: SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding
- Blog: Locking Down the Finetuned LLMs Safety
- Blog: Robust Harmful Features Under Jailbreak Attacks: Mechanistic Evidence from Attention Head Specialization in Large Language Models
- Blog: MetaDefense: Defending Fine-tuning based Jailbreak Attack Before and During Generation
- Blog: SEA: Low-Resource Safety Alignment for Multimodal Large Language Models via Synthetic Embeddings
- Blog: Efficient LLM Moderation with Multi-Layer Latent Prototypes
- Blog: Adversarial Manipulation of Reasoning Models using Internal Representations
- Blog: SAID: Empowering Large Language Models with Self-Activating Internal Defense
- Blog: Leashing the Inner Demons: Self-Detoxification for Language Models
- Blog: Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection
- Blog: Large Language Models can Become Strong Self-Detoxifiers
- Blog: COSMIC: Generalized Refusal Direction Identification in LLM Activations
- Blog: HiddenDetect: Detecting Jailbreak Attacks against Multimodal Large Language Models via Monitoring Hidden States
- Blog: Alleviating the Fear of Losing Alignment in LLM Fine-tuning
- Blog: Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth
- Blog: Intrinsic Guardrails: How Semantic Geometry of Personality Interacts with Emergent Misalignment in LLMs
- Blog: AdaSteer: Your Aligned LLM is Inherently an Adaptive Jailbreak Defender
- Blog: On Prompt-Driven Safeguarding for Large Language Models
- Blog: Minionese: Comprehensive Benchmark and Mechanistic Study of Multilingual LLM Safety
- Blog: Read the Scene, Not the Script: Outcome-Aware Safety for LLMs
- Blog: CoT is Not the Chain of Truth: An Empirical Internal Analysis of Reasoning LLMs for Fake News Generation
History & citation
Current · Permanent link
- This direction split from Latent Space Monitoring for LLM Safety (55% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_latent_safety_awareness_192f46bc,
title = {latent safety awareness},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/latent-safety-awareness/rdr_192f46bc357547318d8137648654cc1b},
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_latent_safety_awareness_192f46bc_v1,
title = {latent safety awareness},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/latent-safety-awareness/rdr_192f46bc357547318d8137648654cc1b/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}