pith. sign in

False sense of security: Why probing- based malicious input detection fails to generalize

3 Pith papers cite this work. Polarity classification is still indexing.

3 Pith papers citing it

years

2026 3

verdicts

UNVERDICTED 3

clear filters

representative citing papers

PRISM: Recovering Instruction Sets from Language Model Activations

cs.AI · 2026-06-08 · unverdicted · novelty 7.0

PRISM is a new activation-conditioned model that recovers full sets of simultaneous instructions from LLM hidden states via judge-guided GRPO training and outperforms prior activation-to-language methods on security-relevant tasks.

citing papers explorer

Showing 1 of 1 citing paper after filters.

  • PRISM: Recovering Instruction Sets from Language Model Activations cs.AI · 2026-06-08 · unverdicted · none · ref 48

    PRISM is a new activation-conditioned model that recovers full sets of simultaneous instructions from LLM hidden states via judge-guided GRPO training and outperforms prior activation-to-language methods on security-relevant tasks.