 Efficient Model Editing with Task-Localized Sparse Fine-tuningInternational Conference on Learning Representations (ICLR), 2025 |
 ObscuraCoder: Powering Efficient Code LM Pre-Training Via Obfuscation GroundingInternational Conference on Learning Representations (ICLR), 2025 |
 CogSteer: Cognition-Inspired Selective Layer Intervention for Efficiently Steering Large Language ModelsAnnual Meeting of the Association for Computational Linguistics (ACL), 2024 |
 Towards Modular LLMs by Building and Reusing a Library of LoRAsInternational Conference on Machine Learning (ICML), 2024 |
 Zero-Shot Tokenizer TransferNeural Information Processing Systems (NeurIPS), 2024 |
