Supply chain Research

Security research involving model, dataset, and alignment supply chains

2 matching entries out of 9 in this category

Supply chain research results

2 entries

Backdoored PEFT adapters and limits of weight-only detection

Published 8/3/2026

Analyzed 9/9/2026

Third-party fine-tuning adapters may contain backdoors. Z-PEFT screens adapter weights using spectral features, evaluated on PADBench's 13,300 adapters.

Z-PEFT: Zero-shot Backdoor Detection in Parameter-Efficient Fine-Tuning via Canonical Spectral Signatures

Evaluated models: Llama 2 7B, Llama 2 13B, Qwen 1.5 7B +2 more

Source: arXiv

RLHF Preference Data Poisoning

Published 3/1/2025

Analyzed 12/9/2025

An untrusted reinforcement-learning-from-human-feedback (RLHF) platform can selectively manipulate preference samples associated with an attacker's target domain. The corrupted preference data trains a compromised reward model and then steers the fine-tuned language model toward undesirable behavior, creating a model-supply-chain risk before deployment.

LLM Misalignment via Adversarial RLHF Platforms

Evaluated models: Not reported

Source: arXiv

Related research categories