AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

AlienLM: Alienization of Language for API-Boundary Privacy in Black-Box LLMs

Jaehee Kim, Pilsung Kang

ICML 2026 · Large language models › Instruction and fine-tuning · approach: Safety alignment after fine-tuning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Alien Adaptation Training (AAT) on top of the vocabulary bijection (ablation)AlienLM · 7-benchmark averageaverage accuracy: 24.77 -> 52.92Improved
use token-ID relabeling instead of character-level ROT13-ASCII substitution (ablation)AlienLM · 7-benchmark averageaverage accuracy: 23.34 -> 52.92Improved
optimize the bijection with embedding similarity instead of random bijection (ablation)AlienLM · 7-benchmark averageaverage accuracy: 36.96 -> 52.92Improved
optimize the bijection with embedding similarity instead of random bijection (ablation)AlienLM · GSM8Kexact match: 31.08 -> 63.08Improved
use proxy-model embeddings rather than target-model embeddings for bijection optimization (ablation)AlienLM · 7-benchmark averageaverage accuracy: 55.07 -> 53.32No clear effect
lower alienization ratio rho (permute fewer tokens) (ablation)AlienLM · 7-benchmark averageaverage accuracy: Pearson r=-0.96; about 86% RR at rho=0.6Improved
exclude domain data from AAT (ablation)AlienLM · GSM8Kaccuracy: 55.5% -> 41.7%Hurt
exclude domain data only from bijection optimization (-Tokenizer) (ablation)AlienLM · General Avg / GSM8KaccuracyNo clear effect
add 150K extra domain examples to AAT (ablation)AlienLM · GSM8Kaccuracy: 55.5% -> 60.6%Improved
train AAT exclusively on domain data (ablation)AlienLM · general benchmarksGeneral Avg: 50.55 -> 44.15Hurt
use per-user bijection seeds instead of a shared seed under collaborative leakage (ablation)AlienLM · Collaborative token recovery (1K pairs)token recovery: 6.05% -> 4.40%Improved
train a single tenant model on five mixed bijection seeds instead of per-key models (ablation)AlienLM · 7-benchmark averageaverage accuracy: 50.95 -> 41.26Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy