What Are Stablecoins and Why Do We Need Them? The Volatility Problem Bitcoin can gain or lose 10% of its value in a single day. Ethereum swings wildly with market sentiment. This volatility makes cryptocurrencies excellent for speculation but terrible for: Daily transactions Store of value Salary payments Savings accounts Anything requiring price stability The problem: If you’re paid in Bitcoin on Monday, by Friday your salary could be worth 20% less—or 20% more. That’s not money, that’s […]

Ver mais

Inspired by Karpathy’s MicroGPT, I wanted to build the equivalent for text diffusion — a minimal implementation that shows the core algorithm without the complexity. Autoregressive models generate left to right. Diffusion generates all tokens at once by iteratively unmasking from noise: _ _ _ _ _ _ → _ o r _ a → n o r i a Three implementations included: – train_minimal.py (143 lines, pure NumPy) — bare minimum – train_pure.py (292 lines, pure NumPy) […]

Ver mais

arXiv:2511.11817v2 Announce Type: replace Abstract: Time series forecasting is essential in a wide range of real world applications. Recently, frequency-domain methods have attracted increasing interest for their ability to capture global dependencies. However, when applied to non-stationary time series, these methods encounter the $textit{spectral entanglement}$ and the computational burden of complex-valued learning. The $textit{spectral entanglement}$ refers to the overlap of trends, periodicities, and noise across the spectrum due to $textit{spectral leakage}$ and the presence of non-stationarity. However, existing […]

Ver mais

arXiv:2603.15684v1 Announce Type: new Abstract: Safety alignment in large language models is typically evaluated under isolated queries, yet real-world use is inherently multi-turn. Although multi-turn jailbreaks are empirically effective, the structure of conversational safety failure remains insufficiently understood. In this work, we study safety failures from a state-space perspective and show that many multi-turn failures arise from structured contextual state evolution rather than isolated prompt vulnerabilities. We introduce STAR, a state-oriented diagnostic framework that treats dialogue history as […]

Ver mais

arXiv:2602.04035v1 Announce Type: new Abstract: As data-intensive applications increasingly strain conventional computing systems, processing-in-memory (PIM) has emerged as a promising paradigm to alleviate the memory wall by minimizing data transfer between memory and processing units. This review presents the recent advances in both stateful and non-stateful logic techniques for PIM, focusing on emerging nonvolatile memory technologies such as resistive random-access memory (RRAM), phase-change memory (PCM), and magnetoresistive random-access memory (MRAM). Both experimentally demonstrated and simulated logic designs are […]

Ver mais

arXiv:2602.10512v1 Announce Type: cross Abstract: We develop a theoretical analysis of LLM-guided formal theorem proving in interactive proof assistants (e.g., Lean) by modeling tactic proposal as a stochastic policy in a finite-horizon deterministic MDP. To capture modern representation learning, we treat the state and action spaces as general compact metric spaces and assume Lipschitz policies. To explain the gap between worst-case hardness and empirical success, we introduce problem distributions generated by a reference policy $q$, including a latent-variable […]

Ver mais

arXiv:2603.02649v1 Announce Type: cross Abstract: Adam and AdamW are a class of default optimizers for training deep learning models in machine learning. These adaptive algorithms converge faster but generalize worse compared to SGD. In fact, their proved generalization error $O(frac{1}{sqrt{N}})$ also is larger than $O(frac{1}{N})$ of SGD, where $N$ denotes training sample size. Recently, although some variants of Adam have been proposed to improve its generalization, their improved generalizations are still unexplored in theory. To fill this gap, […]

Ver mais

I’m working on an optimization problem with a complex environment. Environment is complex in inner working but has only one action input. The action can be either binary, or discrete, or continuous. If the environment is optimized on binary action the maximum reward will be less than when on discrete or continuous actions. PPO works when action is binary or discrete but not when it’s continuous. The input to the model needs to be a value between 0 […]

Ver mais