AI Paper Atlas

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

Hide and Seek in Embedding Space: Geometry-based Steganography and Detection in Large Language Models

Charles Westphal, Keivan Navaie, Fernando Rosas

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
replace parity bucketing with embedding hyperplane bucketing (ASCII encoding) (ablation)Ministral-8B LoRA · In-distribution steganography, TrojanStego-style data (T)exact match rate (%): 24 -> 43Improved
replace parity bucketing with embedding bucketing (ASCII encoding) (ablation)Llama-3.1-8B LoRA · In-distribution steganography, TrojanStego-style data (T)exact match rate (%): 17 -> 30Improved
replace parity bucketing with embedding bucketing (ASCII encoding) (ablation)Llama-3.1-8B LoRA · In-distribution steganography, Wiki prompts (W)exact match rate (%): 38 -> 48Improved
replace parity bucketing with embedding bucketing (ASCII encoding) (ablation)Ministral-8B LoRA · In-distribution steganography, Wiki prompts (W)exact match rate (%): 36 -> 29Hurt
use embedding encoding and embedding bucketing instead of ASCII+parity (ablation)Llama-3.3-70B LoRA · In-distribution steganography, Wiki prompts (W)exact match rate (%): 9 -> 19Improved
use embedding bucketing with ASCII encoding instead of ASCII+parity (ablation)Llama-3.3-70B LoRA · In-distribution steganography, Wiki prompts (W)exact match rate (%): 9 -> 15Improved
use embedding encoding with parity bucketing instead of ASCII+parity (ablation)Llama-3.3-70B LoRA · In-distribution steganography, Wiki prompts (W)exact match rate (%): 9 -> 18Improved
use embedding+embedding instead of ASCII+parity (ablation)Ministral-8B full fine-tune · In-distribution steganography, TrojanStego-style data (T)exact match rate (%): 63 -> 57Hurt
average over 8B models: ASCII+Embedding vs Embedding+Embedding vs Embedding+Parity relative gain over baseline (ablation)8B models · In-distribution steganographyrelative improvement (%): 12% / 33% / 25%Improved
add mixed-motive instruction data (75/25) to Trojan-Stego fine-tuning (ablation)Llama-3.1-8B LoRA ASCII+Embedding · Capability benchmarks (MMLU, ARC-C, HellaSwag, WinoGrande)mean capability change (pp): -9.1 -> -2.7Improved
add mixed-motive instruction data to Trojan-Stego fine-tuning (ablation)Ministral-8B full fine-tune ASCII+Parity · Capability benchmarks (MMLU, ARC-C, HellaSwag, WinoGrande)mean capability change (pp): -31.8 -> -24.9Improved
probe final-layer hidden states of fine-tuned versus base models for bit detectionLlama/Ministral 8B stego models · OOD prompts, linear probe bit accuracybit accuracy (%): 60 -> 95--100Improved

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