The Geometry of Ignorance: Detecting Suppressed Knowledge in LLMs through Activation Fragmentation
cs.AI
Submitted: 2026-11-03
Updated: 2026-11-03
Terminology
Sources
- CRISP: Persistent Concept Unlearning via Sparse Autoencoders
- Language Models are Few-Shot Learners
- Extracting Training Data from Large Language Models
- Erasing Conceptual Knowledge from Language Models
- Transformer Feed-Forward Layers Are Key-Value Memories
- On Effects of Steering Latent Representation for Large Language Model Unlearning
- A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity
- The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
- Eight Methods to Evaluate Robust Unlearning in LLMs
- TOFU: A Task of Fictitious Unlearning for LLMs
- Locating and Editing Factual Associations in GPT
- Language Models as Knowledge Bases?
- Direct Preference Optimization: Your Language Model is Secretly a Reward Model
- Redirection for Erasing Memory (REM): Towards a universal unlearning method for corrupted data
- Extracting Unlearned Information from LLMs with Activation Steering
- Layer by Layer: Uncovering Hidden Representations in Language Models
- Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs
- Large Language Model Unlearning
- Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning
- Universal and Transferable Adversarial Attacks on Aligned Language Models
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents