arXiv (All)
open access
A Gravitational Interpretation of Safety Reversion under Fine-Tuning
Safety alignment in large language models can degrade during post-training even when neither the data nor the objective is intentionally adversarial. Alignment rebound and reverse dynamics suggest that this degradation may reactivate behavior suppressed during safety alignment. Building on these ide
Conceptio Open Archive
Analyze & explore related research in Conceptio
Connect this record directly into your research workflow or AI agent. Explore citations across open repositories, query full text, and export structured proof bundles.
This document is indexed with metadata only: full text is not available in the archive for this record. Open the official source ↗
Related documents
- Alignment Whack-a-Mole : Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models#1123051
- The Third Position: What the Assistant Axis Cannot See#104454
- Coherence-Based Alignment: A Structural Architecture for Preventing Goal Drift in Agentic AI Systems#726487
- Faithful Dual-constrained Erasure for Robust LLM Safety Alignment#1144246
- Faithful Dual-constrained Erasure for Robust LLM Safety Alignment#1136773
Record · ID 1129317
Retrieved via Conceptio: every document is proof-bundled with source, license, and retrieval metadata.
Retrieved via Conceptio: every document is proof-bundled with source, license, and retrieval metadata.
Bring this data into your agent: Connect via MCP or API · 20 free credits to start · no card required
Read the Docs → · View CLI source on GitHub · Conceptio Open Knowledge Archive
Read the Docs → · View CLI source on GitHub · Conceptio Open Knowledge Archive