digitado

In multi-fidelity optimization, biased approximations of varying costs of the target function are available. This paper studies the problem of optimizing a locally smooth function with a limited budget, where the learner has to make a tradeoff between the cost and the bias of these approximations. We first prove lower bounds for the simple regret under different assumptions on the fidelities, based on a cost-to-bias function. We then present the Kometo algorithm which achieves, with additional logarithmic factors, […]

Ver mais

Hi everyone! I am an independent researcher working on Reviser, a language model that generates through cursor-relative edit actions on a mutable canvas. It is autoregressive over edit-history actions rather than final text order, which lets it revise its response while keeping decoding efficiency close to standard autoregressive transformers. My goal is to submit the paper to a conference such as ACL, EMNLP, ICML, or a similar venue, and I would really appreciate technical feedback on things like: […]

Ver mais

How are you, hacker? 🪐 What’s happening in tech today, April 17, 2026? The HackerNoon Newsletter brings the HackerNoon homepage straight to your inbox. On this day, we present you with these top quality stories. From The Engineer in the Machine: Neo Wants to Be a Kaggle Grandmaster in a Box to Veritasium Stole $10,000 From MKBHDs Locked iPhone. Apple and Visa Knew About the Bug for 5 Years., let’s dive right in. Veritasium Stole $10,000 From MKBHDs […]

Ver mais

The Demonstration Veritasium’s Derek Muller published a video this week showing cybersecurity researchers draining $10,000 from MKBHD’s fully locked iPhone using off-the-shelf NFC hardware, a laptop, and a Python script. No passcode entered. No Face ID prompt. No unlock of any kind. The phone sat on a reader, the screen stayed dark, and $10,000 moved. https://www.youtube.com/watch?v=PPJ6NJkmDAo&embedable=true The researchers — Professors Ioana Boureanu from the University of Surrey and Tom Chothia from the University of Birmingham — walked through […]

Ver mais

The gap between AI insiders and everyone else is widening, and the spending, suspicion, and even new vocabulary are starting to show it. While OpenAI is busy buying up everything from finance apps to talk shows, a certain shoe company just rebranded as an AI infrastructure play, and Anthropic unveiled a model it says is too powerful to release publicly …but apparently not too […]

Ver mais

Adapter-based methods have become a cost-effective approach to continual learning (CL) for Large Language Models (LLMs), by sequentially learning a low-rank update matrix for each task. To mitigate catastrophic forgetting, state-of-the-art approaches impose constraints on new adapters with respect to the previous ones, by targeting either subspace or coordinate-wise interference. In this paper, we propose JumpLoRA, a novel framework to adaptively induce sparsity in the Low-Rank Adaptation (LoRA) blocks through the use of JumpReLU gating. The method achieves […]

Ver mais

Adapter-based methods have become a cost-effective approach to continual learning (CL) for Large Language Models (LLMs), by sequentially learning a low-rank update matrix for each task. To mitigate catastrophic forgetting, state-of-the-art approaches impose constraints on new adapters with respect to the previous ones, by targeting either subspace or coordinate-wise interference. In this paper, we propose JumpLoRA, a novel framework to adaptively induce sparsity in the Low-Rank Adaptation (LoRA) blocks through the use of JumpReLU gating. The method achieves […]

Ver mais