This intervention significantly improves the performance of LLaMA models on the TruthfulQA benchmark.

We present a non-linear AI-model designed to reconstruct monthly mean anomalies of the European temperature and precipitation based on the Euro-Atlantic Weather regimes (WR) indices. WR represent recurrent, quasi-stationary, and persistent states of the atmospheric circulation that exert considerable influence over the European weather, therefore offering an opportunity for sub-seasonal to seasonal forecasting. While much research has focused on studying the correlation and impacts of the WR on European weather, the estimation of ground-level climate variables, such as temperature and precipitation, from Euro-Atlantic WR remains largely unexplored and is currently limited to linear methods. The presented AI model can capture and introduce complex non-linearities in the relation between the WR indices, describing the state of the Euro-Atlantic atmospheric circulation and the corresponding surface temperature and precipitation anomalies in Europe. We discuss the AI-model performance in reconstructing the monthly mean two-meter temperature and total precipitation anomalies in the European winter and summer, also varying the number of WR used to describe the monthly atmospheric circulation. We assess the impact of errors on the WR indices in the reconstruction and show that a mean absolute relative error below 80% yields improved seasonal reconstruction compared to the ECMWF operational seasonal forecast system, SEAS5. As a demonstration of practical applicability, we evaluate the model using WR indices predicted by SEAS5, finding slightly better or comparable skill relative to the SEAS5 forecast itself. Our findings demonstrate that WR-based anomaly reconstruction, powered by AI tools, offers a promising pathway for sub-seasonal and seasonal forecasting.

Science Archaeology Who built Scandinavia's oldest wooden plank boat? An ancient fingerprint offers clues. Archeologists are closer to solving the Hjortspring Boat's mysteries. Breakthroughs, discoveries, and DIY tips sent every weekday. Archaeologists examining an ancient boat discovered in Denmark over a century ago are getting some help from a clue usually associated with crime scenes .

Can Bike Riders and Self-Driving Cars Be Friends? Some cycling advocates are on board with robotaxis. Others see the self-driving car boom as perpetuating auto dependency. Los Angeles is a car city, and it's rarely more obvious than from a vulnerable perch on top of a bicycle . Among big cities in the US, LA has a middling-to-bad reputation for bike riding.

Large Language Models increasingly possess capabilities that carry dual-use risks. While data filtering has emerged as a pretraining-time mitigation, it faces significant challenges: labeling whether data is harmful is expensive at scale, and given improving sample efficiency with larger models, even small amounts of mislabeled content could give rise to dangerous capabilities. To address risks associated with mislabeled harmful content, prior work proposed Gradient Routing (Cloud et al., 2024) -- a technique that localizes target knowledge into a dedicated subset of model parameters so they can later be removed. We explore an improved variant of Gradient Routing, which we call Selective GradienT Masking (SGTM), with particular focus on evaluating its robustness to label noise. SGTM zero-masks selected gradients such that target domain examples only update their dedicated parameters. We test SGTM's effectiveness in two applications: removing knowledge of one language from a model trained on a bilingual synthetic dataset, and removing biology knowledge from a model trained on English Wikipedia. In both cases SGTM provides better retain/forget trade-off in the presence of labeling errors compared to both data filtering and a previously proposed instantiation of Gradient Routing. Unlike shallow unlearning approaches that can be quickly undone through fine-tuning, SGTM exhibits strong robustness to adversarial fine-tuning, requiring seven times more fine-tuning steps to reach baseline performance on the forget set compared to a finetuning-based unlearning method (RMU). Our results suggest SGTM provides a promising pretraining-time complement to existing safety mitigations, particularly in settings where label noise is unavoidable.

While deep learning has achieved impressive performance in time series forecasting, it becomes increasingly crucial to understand its decision-making process for building trust in high-stakes scenarios. Existing interpretable models often provide only local and partial explanations, lacking the capability to reveal how heterogeneous and interacting input variables jointly shape the overall temporal patterns in the forecast curve. We propose ProtoTS, a novel interpretable forecasting framework that achieves both high accuracy and transparent decision-making through modeling prototypical temporal patterns. ProtoTS computes instance-prototype similarity based on a denoised representation that preserves abundant heterogeneous information. The prototypes are organized hierarchically to capture global temporal patterns with coarse prototypes while capturing finer-grained local variations with detailed prototypes, enabling expert steering and multi-level interpretability. Experiments on multiple realistic benchmarks, including a newly released LOF dataset, show that ProtoTS not only exceeds existing methods in forecast accuracy but also delivers expert-steerable interpretations for better model understanding and decision support. Time series forecasting has been widely applied in high-stakes scenarios such as load forecasting (Jiang et al., 2024; Y ang et al., 2023), energy management (Deb et al., 2017; Weron, 2014), weather prediction (Angryk et al., 2020; Karevan & Suykens, 2020), all of which involve considerable financial impacts. In these applications, while achieving high forecast accuracy is crucial, understanding why and how the model makes specific predictions is equally important. It aids in preventing substantial financial losses and building the trust necessary (Rojat et al., 2021). A range of explainable time series forecasting methods have been developed to simultaneously ensure interpretability and good predictive performance (Oreshkin et al., 2019; Lim et al., 2021; Zhao et al., 2024; Lin et al., 2024). However, their overall interpretability and potential for further performance improvement are limited, since they mainly provide local, partial explanations for both the output and input sides: C1: For the output side, existing methods (Lim et al., 2021; Zhao et al., 2024) mainly explain the prediction at individual time steps, lacking the ability to help users quickly interpret the reasons behind the overall trend in the forecast curve. For each instance, model computes its similarity to all prototypes to form a prediction, enabling detailed local interpretation.

Cross-lingual information retrieval (CLIR) enables access to multilingual knowledge but remains challenging due to disparities in resources, scripts, and weak cross-lingual semantic alignment in embedding models. Existing pipelines often rely on translation and monolingual retrieval heuristics, which add computational overhead and noise, degrading performance. This work systematically evaluates four intervention types, namely document translation, multilingual dense retrieval with pretrained encoders, contrastive learning at word, phrase, and query-document levels, and cross-encoder re-ranking, across three benchmark datasets. We find that dense retrieval models trained specifically for CLIR consistently outperform lexical matching methods and derive little benefit from document translation. Contrastive learning mitigates language biases and yields substantial improvements for encoders with weak initial alignment, and re-ranking can be effective, but depends on the quality of the cross-encoder training data. Although high-resource languages still dominate overall performance, gains over lexical and document-translated baselines are most pronounced for low-resource and cross-script pairs. These findings indicate that cross-lingual search systems should prioritise semantic multilingual embeddings and targeted learning-based alignment over translation-based pipelines, particularly for cross-script and under-resourced languages.

In Ukraine's'kill-zone', robots are a lifeline to troops trapped on perilous eastern front The toy is delivered, a Ukrainian soldier whispers into the radio. In the dead of night, he and his partner move quickly to roll out their cargo from a van. Speed is crucial as they are within the range of deadly Russian drones. The fifth brigade's new toy is an unmanned ground vehicle (UGV), a robot that provides a lifeline for Ukrainian troops at the front in Pokrovsk and Myrnograd, a strategic hub in eastern Ukraine. Russian forces are relentlessly trying to cut off Ukraine's supply routes in the area.

Text-to-Image (T2I) has witnessed significant advancements, demonstrating superior performance for various generative tasks.