CORTEXA
← Browse
arxivcs.LGcs.CR2026-07-06

Learning Only What Valid Adapters Can Express: Subspace-Constrained Adaptation Against Fine-Tuning Poisoning

Fabien Polly

Parameter-efficient fine-tuning still leaves a broad space of behavior-changing updates reachable, so a poisoned objective can be represented and optimized. We study an alternative: adaptation constrained to the subspace estimated from a trusted pool of existing task adapters. On flan-t5-large with 196 public LoRA adapters, we show that (1) the functionally relevant content of an adapter lies in a low-dimensional shared subspace, 30 to 38 percent of its weight norm being redundant under the evaluated task distributions; (2) gradient adaptation restricted to 128 coordinates on this subspace matches full LoRA fine-tuning on clean classification data, while under targeted label inversion LoRA collapses to 3-26 percent exact match and the constrained learner keeps 62-96 percent on the tasks the pool covers; (3) the constrained learner cannot fit corrupted data, its adaptation loss separating clean from garbage by two orders of magnitude (120x), an out-of-distribution signal without an extra detector; and (4) against an adaptive backdoor attacker who optimizes within the subspace, the attack is blocked (8 percent success versus 100 for LoRA) on the task where its target behavior is unlike anything in the pool, and only partially blocked (85 percent) when the target coincides with a common pool behavior. On these two tasks the outcome is consistent with how close the target is to the pool's directions, which suggests but does not establish a pool-relative boundary. The mechanism trades peak plasticity for these properties: on tasks the pool covers poorly, unconstrained fine-tuning wins, and the protection assumes the pool itself is trusted. Code and data are public.

View free PDFSource page

Related papers

arxivcs.CRcs.AIcs.CYcs.LG2026-07-20

Adversarial Robustness of Phishing Email Detection: A Comparative Study of TF-IDF + Logistic Regression and Fine-Tuned DistilBERT

Tanveer Ahmed, Seyedali Pourmoafil

Phishing emails remain one of the most persistent cybersecurity threats, and machine-learning classifiers are widely used to detect them. Most reported detection accuracies, however, are measured on clean, in-distribution test data rather than on emails deliberately altered to ev…

View free PDFSource page
arxivcs.LGcs.AIcs.CR2026-07-02

kNNGuard: Turning LLM Hidden Activations into a Training-Free Configurable Guardrail

Mahmoud Abdelfattah, Hamid Nasiri, Peter Garraghan

Large language models (LLMs) are increasingly deployed in domains requiring guardrails to detect unsafe, off-topic, or adversarial prompts. Existing guardrails predominantly rely on fine-tuning to build classifiers, which often suffer from low generalization and high inference la…

View free PDFSource page
arxivcs.CVcs.CRcs.LG2026-07-14

Delving into the Temporal Challenges of Unified Video Protection Against Image-to-Video and Fine-Tuning-based Customization

Yuxin Huang, Ziming Hong, Mingming Gong, Wanyu Wang, Jing Zhang, Tongliang Liu

Recent diffusion-based video generation models have enabled high-quality personalized video customization through both tuning-based pipelines, which fine-tune a video diffusion model, and reference-based pipelines such as image-to-video generation. However, these capabilities rai…

View free PDFSource page
arxivcs.LGcs.AIcs.CRcs.CY2026-07-07

Recovering Clinical Utility Under Differential Privacy: Empirical Validation of Adaptive Federated Aggregation on Heterogeneous Cardiovascular Datasets

Rodrigo Tertulino, Laercio Alencar, Ricardo Almeida

Validating federated learning frameworks on real clinical data is an essential step between proof-of-concept demonstrations in controlled synthetic environments and deployment in real multicenter healthcare settings. A prior architectural study by the same authors (Tertulino and…

View free PDFSource page
arxivcs.CRcs.AIcs.IRcs.LG2026-07-09

Beware What You Autocomplete: Forensic Attribution of Backdoored Code Completions

Anjun Gao, Yueyang Quan, Zhuqing Liu, Minghong Fang

Large language models have enabled powerful code completion systems that assist developers by predicting subsequent lines of code. However, these models remain vulnerable to backdoor attacks, where malicious fine-tuning data covertly implants unsafe behaviors. Despite advances in…

View free PDFSource page