How the US left Ukraine exposed to Russia's winter war Will Europe use frozen Russian assets to fund war? How can Ukraine rebuild China ties? Russian forces launched 448 attacks on 34 settlements in Ukraine's front-line Zaporizhia region in a single day, injuring a six-year-old child and damaging homes, cars and other infrastructure, regional governor Ivan Fedorov wrote on the Telegram app. Russian drone, missile and artillery attacks on Ukraine's Kherson region injured five people and damaged homes, including seven high-rise buildings, the local military administration said on Telegram. Russian attacks also continued in Ukraine's Dnipropetrovsk and Sumy regions, but local officials there noted that "fortunately, no people were injured".

Can Ukraine restore its pre-war borders? Why are Tomahawk missiles for Ukraine a'red line' for Russia? Is Russia testing NATO with aerial incursions in Europe? Lithuania, a small Baltic state bordering Belarus and Russia's Kaliningrad, is adapting to new tensions between NATO and Moscow. A member of the Lithuanian Riflemen's Union takes part in a military exercise in central Lithuania [Nils Adler/Al Jazeera] Two members of the Lithuanian Riflemen's Union take part in a military exercise in central Lithuania [Nils Adler/Al Jazeera] On a nearby building is an illuminated decorative Z, a symbol used to show support for the Russian military's full-scale invasion of Ukraine, which began in February 2022.

Does seeing always mean knowing? Large Vision-Language Models (LVLMs) integrate separately pre-trained vision and language components, often using CLIP-ViT as vision backbone. However, these models frequently encounter a core issue of "cognitive misalignment" between the vision encoder (VE) and the large language model (LLM). Specifically, the VE's representation of visual information may not fully align with LLM's cognitive framework, leading to a mismatch where visual features exceed the language model's interpretive range. To address this, we investigate how variations in VE representations influence LVLM comprehension, especially when the LLM faces VE-Unknown data-images whose ambiguous visual representations challenge the VE's interpretive precision. Accordingly, we construct a multi-granularity landmark dataset and systematically examine the impact of VE-Known and VE-Unknown data on interpretive abilities. Our results show that VE-Unknown data limits LVLM's capacity for accurate understanding, while VE-Known data, rich in distinctive features, helps reduce cognitive misalignment. Building on these insights, we propose Entity-Enhanced Cognitive Alignment (EECA), a method that employs multi-granularity supervision to generate visually enriched, well-aligned tokens that not only integrate within the LLM's embedding space but also align with the LLM's cognitive framework. This alignment markedly enhances LVLM performance in landmark recognition. Our findings underscore the challenges posed by VE-Unknown data and highlight the essential role of cognitive alignment in advancing multimodal systems.

The pretrain+fine-tune paradigm is foundational in deploying large language models (LLMs) across a diverse range of downstream applications. Among these, Low-Rank Adaptation (LoRA) stands out for its parameter-efficient fine-tuning (PEFT), producing numerous off-the-shelf task-specific LoRA adapters. However, this approach requires explicit task intention selection, posing challenges for automatic task sensing and switching during inference with multiple existing LoRA adapters embedded in a single LLM. In this work, we introduce MeteoRA (Multiple-Tasks embedded LoRA), a scalable multi-knowledge LoRA fusion framework designed for LLMs. MeteoRA integrates various LoRA adapters in a Mixture-of-Experts (MoE) style into the base LLM, enabling the model to automatically select the most pertinent adapter based on the task input. This advancement significantly enhances the LLM's capability to handle composite tasks that require different adapters to solve various components of the problem. Our evaluations, featuring the LlaMA2-13B and LlaMA3-8B base models equipped with off-the-shelf 28 LoRA adapters through MeteoRA, demonstrate equivalent performance with the individual adapters. Furthermore, both base models equipped with MeteoRA achieve superior performance in sequentially solving composite tasks with ten problems in only a single inference process, highlighting the ability of timely intention switching in MeteoRA embedded LLMs.

This thesis is a corpus-based, quantitative, and typological analysis of the functions of Early Slavic participle constructions and their finite competitors ($jegda$-'when'-clauses). The first part leverages detailed linguistic annotation on Early Slavic corpora at the morphosyntactic, dependency, information-structural, and lexical levels to obtain indirect evidence for different potential functions of participle clauses and their main finite competitor and understand the roles of compositionality and default discourse reasoning as explanations for the distribution of participle constructions and $jegda$-clauses in the corpus. The second part uses massively parallel data to analyze typological variation in how languages express the semantic space of English $when$, whose scope encompasses that of Early Slavic participle constructions and $jegda$-clauses. Probabilistic semantic maps are generated and statistical methods (including Kriging, Gaussian Mixture Modelling, precision and recall analysis) are used to induce cross-linguistically salient dimensions from the parallel corpus and to study conceptual variation within the semantic space of the hypothetical concept WHEN.

Russia's top security agency says it has seized dozens of kilos of explosives sent from Ukraine concealed in Orthodox Christian religious icons that had transited through the European Union. The seizure took place on Tuesday, following an inspection of cargo in the northwestern Pskov region near the Latvian border, the Federal Security Service (FSB) said in a statement. There was no immediate comment by Ukraine, which has been fighting off a Russian invasion since February 2022. The FSB said the cargo had passed through Romania, Hungary, Slovakia, Poland, Lithuania and Latvia, and comprised 70 kilos (154 pounds) of home-made explosives and explosive devices "hidden in icons and ready for use". One person was arrested, it continued, adding that it would seek to track down all those involved, including foreigners, who would then face legal proceedings in Russia.

Senior foreign affairs correspondent Greg Palkot reports the latest. Russia on Tuesday is claiming that so far this year, more than 335,000 people have signed up to fight in its military and volunteer units, although a further deployment to Ukraine is not coming, a report says. Reuters, citing Russian state television, quoted Defense Minister Sergei Shoigu telling top generals that there are "no plans for an additional mobilization" and that "the armed forces have the necessary number of military personnel to conduct the special military operation" in Ukraine. "Since the start of the year, more than 335,000 people have entered military service under contract and in volunteer formations," Shoigu reportedly added. "In September alone, more than 50,000 citizens signed contracts."

Russia's defence ministry said that at least three Ukrainian drones were shot down in the latest attempt by Kyiv to attack sites in Moscow. The ministry said early on Tuesday that air defence systems destroyed two drones over the Kaluga and Tver regions, which border the Moscow region. A third drone was shot down closer to the Russian capital, over the Istra district of the Moscow region, it added. Moscow's Mayor Sergei Sobyanin said the "attack drones which targeted Moscow" were destroyed, according to Russia's TASS news agency. Sobyanin said on the Telegram messaging app that there had been "no casualties", according to initial information.

The United States says it has seen notable military progress by Ukraine's forces fighting in the Zaporizhia region over the last 72 hours. Ukrainian troops achieved "some success against that second line of Russian defences", White House National Security Council spokesperson John Kirby said. A recent drone attack on an airport in northwestern Russia's Pskov region was carried out from within Russian territory, Ukraine's intelligence chief Kyrylo Budanov said. The Pskov regional governor said Russian air defence units had "neutralised an unidentified object" spotted flying over the region, just days after a wave of Ukrainian drones destroyed military planes parked at an airfield in the region. The United States says it has seen notable military progress by Ukraine's forces fighting in the Zaporizhia region over the last 72 hours.

Ukrainian officials are generally tight-lipped about attacks inside Russia, says BBC World Affairs correspondent Paul Adams. But it seems that as the campaign gathers pace, officials in Kyiv are more willing to claim them as part of the country's war effort.