Safe at One Loop, Risky at Another: Aligning Safety Across Recurrent Depths in Looped Language Models

Key points

• Since each recurrent depth can be read out independently, a single LoopLM exposes a broader output space across inference depths, raising an important question: whether safety is preserved throughout recurrent computation.1
• We find that attack success can increase at deeper inference depths, the same query can elicit different safety behaviors across depths, and attacks constructed against one depth can transfer to others.1
• Moreover, SFT and preference alignment do not eliminate these safety gaps, motivating an alignment method designed for LoopLMs. We introduce SafeBridge, which combines lightweight depth specific control of shared recurrent layers, selective state bridging, and joint safety supervision across recurrent depths.1
• Our results show that the safety of a LoopLM cannot be inferred from a single recurrent depth, motivating safety alignment across recurrent computation.1

Sources (1)

• [1]Safe at One Loop, Risky at Another: Aligning Safety Across Recurrent Depths in Looped Language Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 10:48 AM

  Looped Language Models (LoopLMs) provide a parameter efficient approach to scaling model capabilities through repeated use of shared parameters across recurrent steps.

  Since each recurrent depth can be read out independently, a single LoopLM exposes a broader output space across inference depths, raising an important question: whether safety is preserved throughout recurrent computation.

Extractive summary: sentences quoted from the sources.