14 papers
From Memorization to Parameter Interference: How Overtraining Experts Harms Model Merging
Stefan Horoi, Guy Wolf, Eugene Belilovsky +1
Modern deep learning is increasingly characterized by the use of open-weight foundation models that can be fine-tuned on specialized datasets. This has led to a proliferation of ex…
Position: agentic AI orchestration should be Bayes-consistent
Theodore Papamarkou, Pierre Alquier, Matthias Bauer +27
LLMs excel at predictive tasks and complex reasoning tasks, but many high-value deployments rely on decisions under uncertainty, for example, which tool to call, which expert to co…
Detoxifying LLMs via Representation Erasure-Based Preference Optimization
Nazanin Mohammadi Sepahvand, Eleni Triantafillou, Hugo Larochelle +3
Large language models (LLMs) trained on webscale data can produce toxic outputs, raising concerns for safe deployment. Prior defenses, based on applications of DPO, NPO, and simila…
SSFL: Discovering Sparse Unified Subnetworks at Initialization for Efficient Federated Learning
Riyasat Ohib, Bishal Thapaliya, Gintare Karolina Dziugaite +3
In this work, we propose Salient Sparse Federated Learning (SSFL), a streamlined approach for sparse federated learning with efficient communication. SSFL identifies a sparse subne…
From Dormant to Deleted: Tamper-Resistant Unlearning Through Weight-Space Regularization
Shoaib Ahmed Siddiqui, Adrian Weller, David Krueger +3
Recent unlearning methods for LLMs are vulnerable to relearning attacks: knowledge believed-to-be-unlearned re-emerges by fine-tuning on a small set of (even seemingly-unrelated) e…
Continual Learning in Vision-Language Models via Aligned Model Merging
Ghada Sokar, Gintare Karolina Dziugaite, Anurag Arnab +3
Continual learning is conventionally tackled through sequential fine-tuning, a process that, while enabling adaptation, inherently favors plasticity over the stability needed to re…