June 2026

Bullbit, a decentralized perpetual futures exchange (Perp DEX), has officially launched $BUBI, marking the platform’s Token Generation Event (TGE) and opening a new phase of development. According to Bullbit, $BUBI has a fixed maximum supply of 4 billion tokens, operating on the Base L2 network under the ERC-20 standard in its initial phase. On the same day, $BUBI will be officially listed on BingX, expanding access to the token for the global investor community. Notably, the airdrop claim […]

Ver mais

Hey everyone, I recently landed a Computer Vision internship after prepping with this checklist I put together. It starts with core math and ML fundamentals, then moves into the specialized CV topics that actually come up in interviews. I compressed it into just 7 days due to time pressure, so it’s very actionable and easy to personalize for your own pace. Sharing it here in case it’s useful for others prepping for ML/CV roles: → https://github.com/David-Magdy/CVIL Would love […]

Ver mais

The composition of training data, governed by the diversity of sources and their mixing strategy, is a cornerstone of Large Language Model (LLM) pre-training. Online Data Mixing (ODM), the technique of adaptively adjusting data mixtures during training, has emerged as a promising direction to improve efficiency. However, existing methods are constrained by their reliance on a singular optimization perspective, which fundamentally overlooks the need for complex LLM pre-training to consider the dynamic data composition from multiple dimensions. To […]

Ver mais

arXiv:2606.21932v1 Announce Type: new Abstract: Remote sensing change detection (CD) from bi-temporal imagery is critical for applications such as urban monitoring, disaster assessment, and environmental management, yet robust localization remains challenging under sparse changes, noisy labels, and appearance variations. In this paper, we propose Context Sampling Attention (CoSA), a lightweight decoder-side refinement module that explicitly leverages bi-temporal feature correlation as a control signal for adaptive change-aware feature enhancement. This differs from conventional attention mechanisms that rely on implicit […]

Ver mais

arXiv:2606.21930v1 Announce Type: new Abstract: As mental health concerns continue to rise globally, social media has emerged as a vital space where individuals seek emotional support. While prior work on personalized emotional support has leveraged seekers’ emotional states, personas, and situational context, these approaches primarily capture the seeker’s current state, overlooking the formative experiences that shape present concerns. In this work, we propose MindTailor, a framework that generates personalized emotional support responses by constructing a case formulation from […]

Ver mais

arXiv:2606.21929v1 Announce Type: new Abstract: Selective ensemble for modern machine learning systems requires choosing promising model candidates under limited evaluation budgets, while downstream tasks often specify only partial preferences over capabilities such as accuracy, robustness, and reasoning. This setting naturally gives rise to a sequential decision problem under partially specified linear preferences. We formalize it as preference-directed multi-objective bandits (PDMOB), where admissible trade-offs are represented by a polyhedral preference cone. Based on this formulation, we introduce Pareto $C$-optimality, […]

Ver mais

arXiv:2606.21926v1 Announce Type: new Abstract: Web accessibility (A11Y), which ensures web content is perceivable and usable for users with disabilities, is a critical requirement for modern web applications. Yet existing tooling overwhelmingly focuses on detecting A11Y violations rather than repairing them. Automated program repair (APR) techniques appear promising for this setting, but our study shows that state-of-the-art APR systems perform poorly when applied to real-world A11Y violations. Unlike conventional sparse-bug scenarios, web A11Y issues often manifest as multiple […]

Ver mais

arXiv:2606.21924v1 Announce Type: new Abstract: Learned indexes have shown attractive space-time trade-offs in main-memory settings, yet a principled I/O cost model for their disk-resident deployments is still missing, which is a prerequisite for index tuning and query optimization. The practically employed page buffer makes the problem even harder: under typical cache policies, many of the logical page references issued by the index are served by the buffer rather than reaching disk, so the effective physical I/O depends jointly […]

Ver mais

A Vision-Language Model (VLM) was fine-tuned using supervised fine-tuning (SFT) for a 10-class classification task. The resulting model achieved approximately 75% F1 score on the evaluation set and was subsequently deployed. To further improve performance, preference data was collected from production for a specific task containing roughly 400 images. For each image: The SFT model’s prediction was compared against a human-reviewed outcome. Preference pairs were constructed using the model prediction as the rejected response and the human-corrected outcome […]

Ver mais