Semi-supervised learning on real-world graphs is frequently challenged by heterophily, where the observed graph is unreliable or label-disassortative. Many existing graph neural networks either rely on a fixed adjacency structure or attempt to handle structural noise through regularization. In this work, we explicitly capture structural uncertainty by modeling a posterior distribution over signed adjacency matrices, allowing each edge to be positive, negative, or absent. We propose a sparse signed message passing network that is naturally robust to edge noise and heterophily, which can be interpreted from a Bayesian perspective. By combining (i) posterior marginalization over signed graph structures with (ii) sparse signed message aggregation, our approach offers a principled way to handle both edge noise and heterophily. Experimental results demonstrate that our method outperforms strong baseline models on heterophilic benchmarks under both synthetic and real-world structural noise. We provide an anonymous repository at: https://anonymous.4open.science/r/SpaM-F2C8

The proliferation of harmful online content--e.g., toxicity, spam, and negative sentiment--demands robust and adaptable moderation systems. However, prevailing moderation systems are centralized and task-specific, offering limited transparency and neglecting diverse user preferences--an approach ill-suited for privacy-sensitive or decentralized environments. We propose a novel framework that leverages in-context learning (ICL) with foundation models to unify the detection of toxicity, spam, and negative sentiment across binary, multi-class, and multi-label settings. Crucially, our approach enables lightweight personalization, allowing users to easily block new categories, unblock existing ones, or extend detection to semantic variations through simple prompt-based interventions--all without model retraining. Extensive experiments on public benchmarks (TextDetox, UCI SMS, SST2) and a new, annotated Mastodon dataset reveal that: (i) foundation models achieve strong cross-task generalization, often matching or surpassing task-specific fine-tuned models; (ii) effective personalization is achievable with as few as one user-provided example or definition; and (iii) augmenting prompts with label definitions or rationales significantly enhances robustness to noisy, real-world data. Our work demonstrates a definitive shift beyond one-size-fits-all moderation, establishing ICL as a practical, privacy-preserving, and highly adaptable pathway for the next generation of user-centric content safety systems. To foster reproducibility and facilitate future research, we publicly release our code on GitHub and the annotated Mastodon dataset on Hugging Face.

x402 enables Hypertext Transfer Protocol (HTTP) services like application programming interfaces (APIs), data feeds, and inference providers to accept cryptocurrency payments for access. As agents increasingly consume these services, discovery becomes critical: which swap interface should an agent trust? Which data provider is the most reliable? We introduce TraceRank, a reputation-weighted ranking algorithm where payment transactions serve as endorsements. TraceRank seeds addresses with precomputed reputation metrics and propagates reputation through payment flows weighted by transaction value and temporal recency. Applied to x402's payment graph, this surfaces services preferred by high-reputation users rather than those with high transaction volume. Our system combines TraceRank with semantic search to respond to natural language queries with high quality results. We argue that reputation propagation resists Sybil attacks by making spam services with many low-reputation payers rank below legitimate services with few high-reputation payers. Ultimately, we aim to construct a search method for x402 enabled services that avoids infrastructure bias and has better performance than purely volume based or semantic methods.

We propose a meta learning framework for detecting anomalies in human language across diverse domains with limited labeled data. Anomalies in language ranging from spam and fake news to hate speech pose a major challenge due to their sparsity and variability. We treat anomaly detection as a few shot binary classification problem and leverage meta-learning to train models that generalize across tasks. Using datasets from domains such as SMS spam, COVID-19 fake news, and hate speech, we evaluate model generalization on unseen tasks with minimal labeled anomalies. Our method combines episodic training with prototypical networks and domain resampling to adapt quickly to new anomaly detection tasks. Empirical results show that our method outperforms strong baselines in F1 and AUC scores. We also release the code and benchmarks to facilitate further research in few-shot text anomaly detection.

John Oliver covered the dangers of AI on his weekly HBO show, calling it "worryingly corrosive" for society. On Last Week Tonight, Oliver said that the "spread of AI generation tools has made it very easy to flood social media sites with cheap, professional-looking, often deeply weird content" using the term AI slop to describe it all. He referred to it as the "newest iteration of spam" with weird images and videos flooding people's feeds, with some people having "absolutely no idea that it isn't real". Oliver said that it was "extremely likely that we are gonna be drowning in this shit for the foreseeable future". With content such as this, "the whole point is to grab your attention" and given how easy it has become to make it, the barrier of entry has been reduced. Meta has not only joined the game with its own tool but it has also tweaked the algorithm meaning that more than a third of content in your feed is now from accounts you don't follow.

We consider the problem of learning sparse additive models, i.e., functions of the form: f(x) =

Large Language Models (LLMs) have demonstrated exceptional performance across diverse tasks, yet their training remains highly resource-intensive and susceptible to critical challenges such as training instability. A predominant source of this instability stems from gradient and loss spikes, which disrupt the learning process, often leading to costly interventions like checkpoint recovery and experiment restarts, further amplifying inefficiencies. This paper presents a comprehensive investigation into gradient spikes observed during LLM training, revealing their prevalence across multiple architectures and datasets. Our analysis shows that these spikes can be up to $1000\times$ larger than typical gradients, substantially deteriorating model performance. To address this issue, we propose Spike-Aware Adam with Momentum Reset SPAM, a novel optimizer designed to counteract gradient spikes through momentum reset and spike-aware gradient clipping. Extensive experiments, including both pre-training and fine-tuning, demonstrate that SPAM consistently surpasses Adam and its variants across various tasks, including (1) LLM pre-training from 60M to 1B, (2) 4-bit LLM pre-training,(3) reinforcement learning, and (4) Time Series Forecasting. Additionally, SPAM facilitates memory-efficient training by enabling sparse momentum, where only a subset of momentum terms are maintained and updated. When operating under memory constraints, SPAM outperforms state-of-the-art memory-efficient optimizers such as GaLore and Adam-Mini. Our work underscores the importance of mitigating gradient spikes in LLM training and introduces an effective optimization strategy that enhances both training stability and resource efficiency at scale. Code is available at https://github.com/TianjinYellow/SPAM-Optimizer.git

Last week, like the Jews of Exodus painting blood on their lintels, hundreds of thousands of Instagram users posted a block of text to their accounts hoping to avoid the plague of artificial intelligence online. "Goodbye Meta AI," the message began, referring to Facebook's parent company, and continued, "I do not give Meta or anyone else permission to use any of my personal data, profile information or photos." Friends of mine posted it; artists I follow posted it; Tom Brady posted it. In their eagerness to combat the encroachment of A.I., all of them seemed to overlook the fact that merely sharing a meme would do nothing to change their legal rights vis-à-vis Meta or any other tech platform. It is, in fact, possible to prevent Meta from training its A.I. models on your personal data.