Multiple Turns, Short Horizons: Measuring Effective-Horizon Compression in Jailbreaks
Multi-turn jailbreaks use feedback from a victim large language model to adapt their queries over several turns.
ProofPaper ↗
Key points
- We define the effective horizon as the shortest remaining query sequence that recovers the same attack progress.
- Empirically, we find that roughly half of the original query sequence is sufficient on average to recover the same attack progress.
- We show that this shorter history can recover the best point reached during an unsuccessful attack and provide a better starting point for continuing the attack.
- We further find that short reward-propagation windows preserve calibrated defense performance close to full-trajectory returns.
Sources (1)
- [1]Multiple Turns, Short Horizons: Measuring Effective-Horizon Compression in JailbreaksarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 04:44 AM
Multi-turn jailbreaks use feedback from a victim large language model to adapt their queries over several turns.
We define the effective horizon as the shortest remaining query sequence that recovers the same attack progress.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 15, 2026openai/codex rust-v0.163.0: 0.163.0
- Oct 15, 2026Code, Test, and Deploy on One Machine with NVIDIA RTX Spark
- Oct 15, 2026137B A6.8B model from Microsoft?
- Oct 14, 2026anthropics/claude-code v2.1.300
- Oct 14, 2026Lexical Persistence Without Semantic Entrenchment in AI Agent Communities
- Oct 14, 2026Epic says its new vibe-coded Epic Games Launcher is faster than Steam