Agent skills, structured packages of procedural knowledge and executable resources that agents dynamically load at inference time, have become a reliable mechanism for augmenting LLM agents. Yet inference-time skill augmentation is fundamentally limited: retrieval noise introduces irrelevant guidance, injected skill content imposes substantial token overhead, and the model never truly acquires the knowledge it merely follows. We ask whether skills can instead be internalized into model parameters, enabling zero-shot autonomous behavior without any runtime skill retrieval. We introduce […]

Ver mais

The transition from a raw dataset to a fine-tuned Large Language Model (LLM) traditionally involves significant infrastructure overhead, including CUDA environment management and high VRAM requirements. Unsloth AI, known for its high-performance training library, has released Unsloth Studio to address these friction points. The Studio is an open-source, no-code local interface designed to streamline the fine-tuning lifecycle for software engineers and AI professionals. By moving beyond a standard Python library into a local Web UI environment, Unsloth allows […]

Ver mais

arXiv:2602.17811v1 Announce Type: new Abstract: A low out-degree orientation directs each edge of an undirected graph with the goal of minimizing the maximum out-degree of a vertex. In the parallel batch-dynamic setting, one can insert or delete batches of edges, and the goal is to process the entire batch in parallel with work per edge similar to that of a single sequential update and with span (or depth) for the entire batch that is polylogarithmic. In this paper […]

Ver mais

Why Shanghai’s medical AI showdown reveals more about system architecture than clinical superiority The headline writes itself: AI defeats human doctors in medical diagnosis, delivering results in under two seconds versus 13 minutes. But beneath this dramatic 390x speed differential lies a far more nuanced story about specialized model architecture, the evolving role of clinical decision support systems, and the critical distinction between diagnostic speed and comprehensive patient care. The Technical Architecture Behind the Victory Shanghai AI Lab’s Gastrointestinal Multimodal […]

Ver mais

Most pseudo-label selection strategies in semi-supervised learning rely on fixed confidence thresholds, implicitly assuming that prediction confidence reliably indicates correctness. In practice, deep networks are often overconfident: high-confidence predictions can still be wrong, while informative low-confidence samples near decision boundaries are discarded. This paper introduces a Confidence-Variance (CoVar) theory framework that provides a principled joint reliability criterion for pseudo-label selection. Starting from the entropy minimization principle, we derive a reliability measure that combines maximum confidence (MC) with residual-class […]

Ver mais

arXiv:2602.12480v1 Announce Type: new Abstract: The proliferation of Transformer models is often constrained by the significant computational and memory bandwidth demands of deployment. To address this, we present MXFormer, a novel, hybrid, weight-stationary Compute-in-Memory (CIM) accelerator that provides high throughput and efficiency for fixed-model inference on large short-sequence Transformers. Our architecture’s foundation is the use of ultra-dense Charge-Trap Transistors (CTTs) in Microscaling MXFP4 CIM arrays, uniquely enabling the on-chip storage of up to hundreds of millions of parameters […]

Ver mais

arXiv:2602.03994v1 Announce Type: new Abstract: Chain-of-thought (CoT) prompting is widely assumed to expose a model’s reasoning process and improve transparency. We attempted to enforce this assumption by penalizing unfaithful reasoning, but found that surface-level compliance does not guarantee causal reliance. Our central finding is negative: even when CoT is verbose, strategic, and flagged by surface-level manipulation detectors, model answers are often causally independent of the CoT content. We present a diagnostic framework for auditing this failure mode: it […]

Ver mais

arXiv:2603.16661v2 Announce Type: replace-cross Abstract: Standard masked discrete diffusion models face limitations in reasoning tasks due to their inability to correct their own mistakes on the masking path. Since they rely on a fixed number of denoising steps, they are unable to adjust their computation to the complexity of a given problem. To address these limitations, we introduce a method based on learning a Markov transition kernel that is trained on its own outputs. This design enables tokens […]

Ver mais

Deep neural networks (DNNs) have been increasingly explored for receiver design because they can handle complex environments without relying on explicit channel models. Nevertheless, because communication channels change rapidly, their distributions can shift over time, often making periodic retraining necessary. This paper proposes a zero-overhead online and continual learning framework for orthogonal frequency-division multiplexing (OFDM) neural receivers that directly detect the soft bits of received signals. Unlike conventional fine-tuning methods that rely on dedicated training intervals or full […]

Ver mais