Cline became ‘agentic’ before it was cool, but building on the bleeding edge usually leads to some structural debt. Over time, the agent loop and the VS Code extension became a package deal—making it a headache to maintain or move to new environments. Its tough to just keep layering features on a rigid core. Cline, the open-source AI coding agent used by millions of developers, shipped a significant architectural change this week: it extracted its internal agent harness […]

Ver mais

Research: JavaScript Sandboxing Research Aaron Harper wrote about Node.js worker threads, which inspired me to run a research task to see if they might help with running JavaScript in a sandbox. Claude Code went way beyond my initial question and produced a comparison of isolated-vm, vm2, quickjs-emscripten, QuickJS-NG, ShadowRealm, and Deno Workers. Tags: sandboxing, javascript, nodejs, claude-code

Ver mais

:::info Authors: Dani Valevski Yaniv Leviathan Moab Arar Shlomi Fruchter ::: Table of Links ABSTRACT 1 INTRODUCTION 2 INTERACTIVE WORLD SIMULATION 3 GAMENGEN 3.1 DATA COLLECTION VIA AGENT PLAY 3.2 TRAINING THE GENERATIVE DIFFUSION MODEL 4 EXPERIMENTAL SETUP 4.1 AGENT TRAINING 4.2 GENERATIVE MODEL TRAINING 5 RESULTS 5.1 SIMULATION QUALITY 5.2 ABLATIONS 6 RELATED WORK 7 DISCUSSION, ACKNOWLEDGEMENTS AND REFERENCES ABSTRACT We present GameNGen, the first game engine powered entirely by a neural model that also enables real-time […]

Ver mais

arXiv:2603.03303v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly used to simulate how specific users respond to a given context, enabling more user-centric applications that rely on user feedback. However, existing user simulators mostly imitate surface-level patterns and language styles, which fail to reflect the underlying states of real users (e.g., beliefs and emotions). To address these limitations, we propose a novel training framework, HumanLM, which builds user simulators that accurately reflect real users. Our key […]

Ver mais

Privacy-Preserving Federated Learning (PPFL) is a Decentralized machine learning paradigm that enables multiple participants to collaboratively train a global model without sharing their data with the integration of cryptographic and privacy-based techniques to enhance the security of the global system. This privacy-oriented approach makes PPFL a highly suitable solution for training shared models in sectors where data privacy is a critical concern. In traditional FL, local models are trained on edge devices, and only model updates are shared […]

Ver mais

We study learning under a two-step contrastive example oracle, as introduced by Mansouri et. al. (2025), where each queried (or sampled) labeled example is paired with an additional contrastive example of opposite label. While Mansouri et al. assume an idealized setting, where the contrastive example is at minimum distance of the originally queried/sampled point, we introduce and analyze a mechanism, parameterized by a non-decreasing noise function $f$, under which this ideal contrastive example is perturbed. The amount of […]

Ver mais

Many reinforcement learning (RL) problems admit multiple terminal solutions of comparable quality, where the goal is not to identify a single optimum but to represent a diverse set of high-quality outcomes. Nevertheless, policies trained by standard expected return maximization routinely collapse onto a small subset of outcomes, a phenomenon commonly attributed to insufficient exploration or weak regularization. We show that this explanation is incomplete: outcome level mode collapse is a structural consequence of the expected-return objective itself. Under […]

Ver mais

Large Reasoning Models (LRMs) have shown exceptional reasoning capabilities, but they also suffer from the issue of overthinking, often generating excessively long and redundant answers. For problems that exceed the model’s capabilities, LRMs tend to exhibit the overconfidence phenomenon, generating overly short but incorrect answers, which may contribute to suboptimal performance. To address these issues, we propose Difficulty-Differentiated Policy Optimization (DDPO), an efficient reinforcement learning algorithm that optimizes simple and complex tasks separately based on the overconfidence phenomenon. […]

Ver mais