In Part 1, a smart account was deployed and the first UserOperation successfully executed through the EntryPoint. At that point, everything worked — but a critical part of the system stayed mostly invisible: the bundler. Bundlers are the bridge between account abstraction and the Ethereum execution layer. They take UserOperations from a separate mempool, pay gas costs upfront, and get reimbursed through the protocol. Understanding how they work—the validation rules, reputation system, and economic incentives—is essential for debugging […]

Ver mais

Temporal Knowledge Graph Question Answering (TKGQA) is inherently challenging, as it requires sophisticated reasoning over dynamic facts with multi-hop dependencies and complex temporal constraints. Existing methods rely on fixed workflows and expensive closed-source APIs, limiting flexibility and scalability. We propose Temp-R1, the first autonomous end-to-end agent for TKGQA trained through reinforcement learning. To address cognitive overload in single-action reasoning, we expand the action space with specialized internal actions alongside external action. To prevent shortcut learning on simple questions, […]

Ver mais

arXiv:2602.00981v1 Announce Type: new Abstract: Spoken question-answering (SQA) systems relying on automatic speech recognition (ASR) often struggle with accurately recognizing medical terminology. To this end, we propose MedSpeak, a novel knowledge graph-aided ASR error correction framework that refines noisy transcripts and improves downstream answer prediction by leveraging both semantic relationships and phonetic information encoded in a medical knowledge graph, together with the reasoning power of LLMs. Comprehensive experimental results on benchmarks demonstrate that MedSpeak significantly improves the accuracy […]

Ver mais

Hey everyone! I recently spent a couple of weekends improving Karpathy’s excellent LLM Council Open Source Project. The original project was brilliant but lacked usability and flexibility imho. What I added: Web search integration (DuckDuckGo, Tavily, Brave, Jina AI) Clean Modern UI with a settings page to support: Support for multiple API providers (OpenRouter, Anthropic, OpenAI, Google, etc.) Customizable system prompts and temperature controls (the custom prompts open up tons of use cases beyond a “council”) Export & […]

Ver mais

arXiv:2601.09849v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed to support human decision-making. This use of LLMs has concerning implications, especially when their prescriptions affect the welfare of others. To gauge how LLMs make social decisions, we explore whether five leading models produce sensible strategies in the repeated prisoner’s dilemma, which is the main metaphor of reciprocal cooperation. First, we measure the propensity of LLMs to cooperate in a neutral setting, without using language reminiscent […]

Ver mais

I finally did it. Last week, I built a prompt that generated a flawless documentation site from a GitHub repo. It was beautiful. I felt like a wizard. I even bookmarked it as my “Gold Standard” prompt. Then, yesterday happened. I ran the exact same prompt on a new repo—similar structure, similar size—and it was a total disaster. The AI started ignoring the CSS requirements, forgot to link the sub-pages, and kept trying to write the docs in […]

Ver mais

In this blog post, I am (briefly) reviewing Christoph Molnar’s *Interpretable Machine Learning Book*. Then, I am writing about two classic generalized linear models, linear and logistic regression. Mainly, this blog post explains the relationship between feature weights and predictions and demonstrates how to construct confidence intervals via Python.

Ver mais

arXiv:2601.21200v2 Announce Type: replace Abstract: Classifier-guided diffusion models generate conditional samples by augmenting the reverse-time score with the gradient of the log-probability predicted by a probabilistic classifier. In practice, this classifier is usually obtained by minimizing an empirical loss function. While existing statistical theory guarantees good generalization performance when the sample size is sufficiently large, it remains unclear whether such training yields an effective guidance mechanism. We study this question in the context of cross-entropy loss, which is […]

Ver mais

In the previous case, I described how we optimized the art pipeline and produced a full set of production artifacts, accelerating art development by approximately 70% (you can read more about it here). However, the majority of these results would not have been achievable without addressing another critical factor—one that is often overlooked, yet fully capable of derailing even the best plans. That factor is the quality of the working day and cross-department interaction. Let’s walk through it […]

Ver mais