5 papers
The Appeal and Reality of Recycling LoRAs with Adaptive Merging
Haokun Liu, Gyung Hyun Je, Marco Ciccone +3
The widespread availability of fine-tuned LoRA modules for open pre-trained models has led to an interest in methods that can adaptively merge LoRAs to improve performance. These m…
TokSuite: Measuring the Impact of Tokenizer Choice on Language Model Behavior
Gül Sena AltıntaÅ, Malikeh Ehghaghi, Brian Lester +4
Tokenizers provide the fundamental basis through which text is represented and processed by language models (LMs). Despite the importance of tokenization, its role in LM performanc…
FlexRank: Nested Low-Rank Knowledge Decomposition for Adaptive Model Deployment
Riccardo Zaccone, Stefanos Laskaridis, Marco Ciccone +1
The growing scale of deep neural networks, encompassing large language models (LLMs) and vision transformers (ViTs), has made training from scratch prohibitively expensive and depl…
Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models
Filippo Rinaldi, Aniello Panariello, Giacomo Salici +4
When a new release of a foundation model is published, practitioners typically need to repeat fine-tuning, even if the same task was already tackled in the previous version. A prom…
LoopFormer: Elastic-Depth Looped Transformers for Latent Reasoning via Shortcut Modulation
Ahmadreza Jeddi, Marco Ciccone, Babak Taati
Looped Transformers have emerged as an efficient and powerful class of models for reasoning in the language domain. Recent studies show that these models achieve strong performance…