Kyiv is many miles from the front line, but Ukraine's war with Russia is never far away - with Moscow's missile and drone attacks directed at the city almost every day. On the frozen surface of the mighty River Dnipro, the BBC speaks to men who spend hours fishing to take their minds off the almost four-year-old conflict, which has left homes with no heating after Russian strikes on power stations. Drilling holes in the ice of the river in the heart of the city, these ice-fisherman - many of them veterans with friends and family at the front - hope to catch small fish, and a little respite. Authorities deliberately triggered the avalanche on Mount Elbrus to release a build up of snow. The limited deployment involves Germany, France, Sweden, Norway, Finland, the Netherlands and the UK.

Archaeologists also located English Civil War cannonballs and a Tudor-era shoe near a Newcastle river. Breakthroughs, discoveries, and DIY tips sent every weekday. At the height of its power, the Roman Empire extended as far away as Britain . Based on a new trove of archaeological artifacts discovered in northeast England, Britain hosted critical sites that supplied the empire's vast military complex. Over six months in 2025, researchers from the United Kingdom's Durham University excavated the new evidence on the banks of the River Wear not far from Newcastle, England.

Existing evaluation protocols for brain visual decoding predominantly rely on coarse metrics that obscure inter-model differences, lack neuroscientific foundation, and fail to capture fine-grained visual distinctions. To address these limitations, we introduce BASIC, a unified, multigranular evaluation framework that jointly quantifies structural fidelity, inferential alignment, and contextual coherence between decoded and ground-truth images. For the structural level, we introduce a hierarchical suite of segmentation-based metrics, including foreground, semantic, instance, and component masks, anchored in granularity-aware correspondence across mask structures. For the semantic level, we extract structured scene representations encompassing objects, attributes, and relationships using multimodal large language models, enabling detailed, scalable, and context-rich comparisons with ground-truth stimuli. We benchmark a diverse set of visual decoding methods across multiple stimulus-neuroimaging datasets within this unified evaluation framework. Together, these criteria provide a more discriminative, interpretable, and comprehensive foundation for evaluating brain visual decoding methods.

This paper presents a mathematical interpretation of self-attention by connecting it to distributional semantics principles. We show that self-attention emerges from projecting corpus-level co-occurrence statistics into sequence context. Starting from the co-occurrence matrix underlying GloVe embeddings, we demonstrate how the projection naturally captures contextual influence, with the query-key-value mechanism arising as the natural asymmetric extension for modeling directional relationships. Positional encodings and multi-head attention then follow as structured refinements of this same projection principle. Our analysis demonstrates that the Transformer architecture's particular algebraic form follows from these projection principles rather than being an arbitrary design choice.

While looped transformers (termed as Looped-Attn) often outperform standard transformers (termed as Single-Attn) on complex reasoning tasks, the theoretical basis for this advantage remains underexplored. In this paper, we explain this phenomenon through the lens of loss landscape geometry, inspired by empirical observations of their distinct dynamics at both sample and Hessian levels. To formalize this, we extend the River-Valley landscape model by distinguishing between U-shaped valleys (flat) and V-shaped valleys (steep). Based on empirical observations, we conjecture that the recursive architecture of Looped-Attn induces a landscape-level inductive bias towards River-V-Valley. Theoretical derivations based on this inductive bias guarantee a better loss convergence along the river due to valley hopping, and further encourage learning about complex patterns compared to the River-U-Valley induced by Single-Attn. Building on this insight, we propose SHIFT (Staged HIerarchical Framework for Progressive Training), a staged training framework that accelerates the training process of Looped-Attn while achieving comparable performances.

Language models (LMs) often struggle to generate diverse, human-like creative content, raising concerns about the long-term homogenization of human thought through repeated exposure to similar outputs. Yet scalable methods for evaluating LM output diversity remain limited, especially beyond narrow tasks such as random number or name generation, or beyond repeated sampling from a single model. We introduce Infinity-Chat, a large-scale dataset of 26K diverse, real-world, open-ended user queries that admit a wide range of plausible answers with no single ground truth. We introduce the first comprehensive taxonomy for characterizing the full spectrum of open-ended prompts posed to LMs, comprising 6 top-level categories (e.g., brainstorm & ideation) that further breaks down to 17 subcategories. Using Infinity-Chat, we present a large-scale study of mode collapse in LMs, revealing a pronounced Artificial Hivemind effect in open-ended generation of LMs, characterized by (1) intra-model repetition, where a single model consistently generates similar responses, and more so (2) inter-model homogeneity, where different models produce strikingly similar outputs. Infinity-Chat also includes 31,250 human annotations, across absolute ratings and pairwise preferences, with 25 independent human annotations per example. This enables studying collective and individual-specific human preferences in response to open-ended queries. Our findings show that LMs, reward models, and LM judges are less well calibrated to human ratings on model generations that elicit differing idiosyncratic annotator preferences, despite maintaining comparable overall quality. Overall, INFINITY-CHAT presents the first large-scale resource for systematically studying real-world open-ended queries to LMs, revealing critical insights to guide future research for mitigating long-term AI safety risks posed by the Artificial Hivemind.

Sea lampreys--invasive, leechlike creatures that once nearly destroyed the Great Lakes' fishing economy--are kept in check by a small U.S.-Canadian program. Will it survive Trump's slash-and-burn campaign? Ally Porter walked ahead of me as we sidestepped down a steep, loamy embankment. Our path lit only by headlamps, a waning sliver of moon, and what seemed to be thousands of stars, we made our way to a mucky riverbank about twenty feet below. At one point, I lost my footing and ended up wedged against a tree trunk. Porter, who had two tight braids that landed just below her shoulders, kept going. She moved with ease through several inches of sludge, toward a yellow glow stick tied to a tree at the water's edge.

While recent advances in reasoning models have demonstrated cognitive behaviors through reinforcement learning, existing approaches struggle to invoke deep reasoning capabilities in multi-turn agents with long-horizon interactions. We propose DeepMiner, a novel framework that elicits such abilities by introducing high-difficulty training tasks and dynamic context window. DeepMiner presents a reverse construction method to generate complex but verifiable question-answer pairs from authentic web sources, which ensures the challenge and reliability of training data while injecting cognitive capabilities into multi-turn reasoning scenarios. We further design an elegant yet effective dynamic context management strategy for both training and inference, utilizing sliding window mechanisms while eliminating the dependency on external summarization models, thereby efficiently empowering the model to handle continuously expanding long-horizon contexts. Through reinforcement learning on Qwen3-32B, we develop DeepMiner-32B, which achieves substantial performance improvements across multiple search agent benchmarks. DeepMiner attains 33.5% accuracy on BrowseComp-en, surpassing the previous best open-source agent by almost 20 percentage points, and demonstrates consistent improvements on BrowseComp-zh, XBench-DeepSearch, and GAIA. Notably, our dynamic context management enables sustained interactions of nearly 100 turns within standard 32k context length, effectively addressing the context limitations that constrain existing multi-turn interaction systems.