MIT Libraries: Recent submissions

Now showing items 19-21 of 64589

    • Unforgettable Generalization in Language Models 

      Zhang, Eric (Massachusetts Institute of Technology, 2025-05)
      When language models (LMs) are trained to forget (or “unlearn”) a skill, how precisely does their behavior change? We study the behavior of transformer LMs in which tasks have been forgotten via fine-tuning on randomized ...

    • Guessing Random Additive Noise Decoding in Coded Multiple-Input Multiple-Output Systems 

      Wu, Benjamin (Massachusetts Institute of Technology, 2025-05)
      Multiple-Input Multiple-Output (MIMO) wireless communication systems incorporate forward error correction (FEC) to achieve high reliability under fading and interference. In this thesis, we explore the emerging FEC paradigm ...

    • Improving Accuracy Predictions of Companion Classifiers for LLM Routing 

      Wu, Jessica L. (Massachusetts Institute of Technology, 2025-05)
      The increasing versatility of Large Language Models (LLMs) calls for developing effective routing systems to match tasks with the most suitable models, balancing accuracy and computational cost. This research introduces a ...