The digital exclusion of endangered languages remains a critical challenge in NLP, limiting both linguistic research and revitalization efforts. This study introduces the first computational investigation of Comanche, an Uto-Aztecan language on the verge of extinction, demonstrating how minimal-cost, community-informed NLP interventions can support language preservation. We present a manually curated dataset of 412 phrases, a synthetic data generation pipeline, and an empirical evaluation of GPT-4o and GPT-4o-mini for language identification. Our experiments reveal that while LLMs struggle with Comanche in zero-shot settings, few-shot prompting significantly improves performance, achieving near-perfect accuracy with just five examples. Our findings highlight the potential of targeted NLP methodologies in low-resource contexts and emphasize that visibility is the first step toward inclusion. By establishing a foundation for Comanche in NLP, we advocate for computational approaches that prioritize accessibility, cultural sensitivity, and community engagement.

Human evaluation of generated language through pairwise preference judgments is pervasive. However, under common scenarios, such as when generations from a model pair are very similar, or when stochastic decoding results in large variations in generations, it results in inconsistent preference ratings. We address these challenges by introducing a meta-evaluation measure, separability, which estimates how suitable a test instance is for pairwise preference evaluation. For a candidate test instance, separability samples multiple generations from a pair of models, and measures how distinguishable the two sets of generations are. Our experiments show that instances with high separability values yield more consistent preference ratings from both human- and auto-raters. Further, the distribution of separability allows insights into which test benchmarks are more valuable for comparing models. Finally, we incorporate separability into ELO ratings, accounting for how suitable each test instance might be for reliably ranking LLMs. Overall, separability has implications for consistent, efficient and robust preference evaluation of LLMs with both human- and auto-raters.

Misinformation is a prevalent societal issue due to its potential high risks. Out-of-context (OOC) misinformation, where authentic images are repurposed with false text, is one of the easiest and most effective ways to mislead audiences. Current methods focus on assessing image-text consistency but lack convincing explanations for their judgments, which is essential for debunking misinformation. While Multimodal Large Language Models (MLLMs) have rich knowledge and innate capability for visual reasoning and explanation generation, they still lack sophistication in understanding and discovering the subtle crossmodal differences. In this paper, we introduce SNIFFER, a novel multimodal large language model specifically engineered for OOC misinformation detection and explanation. SNIFFER employs two-stage instruction tuning on InstructBLIP. The first stage refines the model's concept alignment of generic objects with news-domain entities and the second stage leverages language-only GPT-4 generated OOC-specific instruction data to fine-tune the model's discriminatory powers. Enhanced by external tools and retrieval, SNIFFER not only detects inconsistencies between text and image but also utilizes external knowledge for contextual verification. Our experiments show that SNIFFER surpasses the original MLLM by over 40% and outperforms state-of-the-art methods in detection accuracy. SNIFFER also provides accurate and persuasive explanations as validated by quantitative and human evaluations.

Recent Reference-Based image super-resolution (RefSR) has improved SOTA deep methods introducing attention mechanisms to enhance low-resolution images by transferring high-resolution textures from a reference high-resolution image. The main idea is to search for matches between patches using LR and Reference image pair in a feature space and merge them using deep architectures. However, existing methods lack the accurate search of textures. They divide images into as many patches as possible, resulting in inefficient memory usage, and cannot manage large images. Herein, we propose a deep search with a more efficient memory usage that reduces significantly the number of image patches and finds the $k$ most relevant texture match for each low-resolution patch over the high-resolution reference patches, resulting in an accurate texture match. We enhance the Super Resolution result adding gradient density information using a simple residual architecture showing competitive metrics results: PSNR and SSMI.

For example, you might say: The purpose of this data collection is to further speech-to-speech translation research by creating an open collection of translated conversations, something that has not been done before. Today you will have a conversation with your partner in one language, then re-enact parts of it in another language. I will select some snippets of the audio and replay 7 them for you to translate re-record in the other language. It is important that you try your best to make it sound natural while also keeping the same feeling as in the original. Try to recreate pauses, laughs, long breaths, or anything of that sort during the second recording if possible. I can replay the audio as many times as you need, and give you as much time as you need to translate. If either of us feel like you can translate the words better or if the prosody was not as faithful in feeling as it could be, then we can redo it until we are satisfied. Please be vocal of any opinions that you have about the process and ask any questions that may arise.

Last year, the United States Department of Homeland Security advertised the impending "deployment" on the US-Mexico border of "robot dogs". According to a celebratory feature article published on the department's website, the goal of the programme was to "force-multiply" the presence of US Customs and Border Protection (CBP) as well as to "reduce human exposure to life-threatening hazards". In case there was any doubt as to which human lives were of concern, the article specified: "The American Southwest is a region that blends a harsh landscape, temperature extremes and various other non-environmental threats that can create dangerous obstacles for those who patrol the border." There is no denying that the US-Mexico border is an inhospitable place; just ask the countless refuge seekers who have died trying to navigate it, thanks in large part to ongoing US efforts to effectively criminalise the very right to asylum. And the terrain is becoming ever more hostile with the mad dash to run the entire world on artificial intelligence, border "security" operations to boot. The proliferation of AI-reliant surveillance technology has increasingly forced undocumented people into ever more dangerous territory, where "non-environmental threats" will apparently now also include canine robots.

Artificial intelligence (AI) and machine learning (ML) are expanding in popularity for broad applications to challenging tasks in chemistry and materials science. Examples include the prediction of properties, the discovery of new reaction pathways, or the design of new molecules. The machine needs to read and write fluently in a chemical language for each of these tasks. Strings are a common tool to represent molecular graphs, and the most popular molecular string representation, SMILES, has powered cheminformatics since the late 1980s. However, in the context of AI and ML in chemistry, SMILES has several shortcomings -- most pertinently, most combinations of symbols lead to invalid results with no valid chemical interpretation. To overcome this issue, a new language for molecules was introduced in 2020 that guarantees 100\% robustness: SELFIES (SELF-referencIng Embedded Strings). SELFIES has since simplified and enabled numerous new applications in chemistry. In this manuscript, we look to the future and discuss molecular string representations, along with their respective opportunities and challenges. We propose 16 concrete Future Projects for robust molecular representations. These involve the extension toward new chemical domains, exciting questions at the interface of AI and robust languages and interpretability for both humans and machines. We hope that these proposals will inspire several follow-up works exploiting the full potential of molecular string representations for the future of AI in chemistry and materials science.

Rep. August Pfluger joins'Fox & Friends First' and calls out Biden's handling of border crisis The feds have turned to cutting-edge cameras developed by a virtual reality wunderkind to help them monitor the southern border -- by creating an invisible border wall. The high-tech watch poles known as Autonomous Surveillance Towers are powered by solar energy and use artificial intelligence to detect movement along a two-mile radius, sending the information in real-time to agents patrolling the area. And they're now being installed at different points along the nearly 2,000 miles of the US-Mexico border. "The ASTs are in remote locations that are difficult to reach," Border Patrol agent Joel Freeland recently told The Post. "They operate 24-hours a day and are environmentally friendly because they rely entirely on solar power." The ASTs were developed by Palmer Luckey, the 28-year-old founder and designer of Oculus VR and Oculus Rift.

Mexico's increasingly militarized crackdown of powerful drug cartels has left nearly 39,000 unidentified bodies languishing in the country's morgues – a grotesque symbol of the ever-burgeoning war on drugs and rampant violence. Investigative NGO Quinto Elemento Labs, in a recent report, found that an alarming number of people have been simply buried in common graves without proper postmortems, while others were left in funeral homes. The so-called war of drugs has claimed the lives of nearly 300,000 people over the last 14 years, while another 73,000 have gone missing. All the while, these cartels have yet to be designated formal terrorist organizations despite boasting well-documented arsenals of sophisticated weaponry to rival most fear-inducing militias on battlefields abroad. Just last month, reports surfaced that Mexico's Jalisco New Generation Cartel (CJNG) now possess bomb-toting drones – which The Drive's Warzone depicts as "small quadcopter-type drones carrying small explosive devices to attack its enemies."

Adversarial data poisoning is an effective attack against machine learning and threatens model integrity by introducing poisoned data into the training dataset. So far, it has been studied mostly for classification, even though regression learning is used in many mission critical systems (such as dosage of medication, control of cyber-physical systems and managing power supply). Therefore, in the present research, we aim to evaluate all aspects of data poisoning attacks on regression learning, exceeding previous work both in terms of breadth and depth. We present realistic scenarios in which data poisoning attacks threaten production systems and introduce a novel black-box attack, which is then applied to a real-word medical use-case. As a result, we observe that the mean squared error (MSE) of the regressor increases to 150 percent due to inserting only two percent of poison samples. Finally, we present a new defense strategy against the novel and previous attacks and evaluate it thoroughly on 26 datasets. As a result of the conducted experiments, we conclude that the proposed defence strategy effectively mitigates the considered attacks.