LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via interventions.
What Features in Prompts Jailbreak LLMs? Investigating the Mechanisms Behind Attacks , shorttitle =
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
Causal mediation analysis shows harmful LLM outputs arise in late layers from MLP failures and gating neurons, with early layers handling harm context detection and signal propagation.
Probes on persona principal components from contrastive prompts generalize better than raw activation probes for harmful behaviors across 10 datasets.
citing papers explorer
-
Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space
LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via interventions.
-
Why Do Large Language Models Generate Harmful Content?
Causal mediation analysis shows harmful LLM outputs arise in late layers from MLP failures and gating neurons, with early layers handling harm context detection and signal propagation.
-
Do Linear Probes Generalize Better in Persona Coordinates?
Probes on persona principal components from contrastive prompts generalize better than raw activation probes for harmful behaviors across 10 datasets.