The fact that everyone with a social media account can create and share content, and the increasing public reliance on social media platforms as a news and information source bring about significant challenges such as misinformation, fake news, harmful content, etc. Although human content moderation may be useful to an extent and used by these platforms to flag posted materials, the use of AI models provides a more sustainable, scalable, and effective way to mitigate these harmful contents. However, low-resourced languages such as the Somali language face limitations in AI automation, including scarce annotated training datasets and lack of language models tailored to their unique linguistic characteristics. This paper presents part of our ongoing research work to bridge some of these gaps for the Somali language. In particular, we created two human-annotated social-media-sourced Somali datasets for two downstream applications, fake news \& toxicity classification, and developed a transformer-based monolingual Somali language model (named SomBERTa) -- the first of its kind to the best of our knowledge. SomBERTa is then fine-tuned and evaluated on toxic content, fake news and news topic classification datasets. Comparative evaluation analysis of the proposed model against related multilingual models (e.g., AfriBERTa, AfroXLMR, etc) demonstrated that SomBERTa consistently outperformed these comparators in both fake news and toxic content classification tasks while achieving the best average accuracy (87.99%) across all tasks. This research contributes to Somali NLP by offering a foundational language model and a replicable framework for other low-resource languages, promoting digital and AI inclusivity and linguistic diversity.

Recent advancements in Large Reasoning Models (LRMs), such as OpenAI's o1/o3 and DeepSeek-R1, have demonstrated remarkable performance in specialized reasoning tasks through human-like deliberative thinking and long chain-of-thought reasoning. However, our systematic evaluation across various model families (DeepSeek, Qwen, and LLaMA) and scales (7B to 671B) reveals that acquiring these deliberative reasoning capabilities significantly reduces the foundational capabilities of LRMs, including notable declines in helpfulness and harmlessness, alongside substantially increased inference costs. Importantly, we demonstrate that adaptive reasoning -- employing modes like Zero-Thinking, Less-Thinking, and Summary-Thinking -- can effectively alleviate these drawbacks. Our empirical insights underline the critical need for developing more versatile LRMs capable of dynamically allocating inference-time compute according to specific task characteristics.

Data scarcity is a long-standing challenge in the Vision-Language Navigation (VLN) field, which extremely hinders the generalization of agents to unseen environments. Previous works primarily rely on additional simulator data or web-collected images/videos to improve the generalization. However, the simulator environments still face limited diversity, and the web-collected data often requires extensive labor to remove the noise. In this paper, we propose a Rewriting-driven AugMentation (RAM) paradigm for VLN, which directly creates the unseen observation-instruction pairs via rewriting human-annotated training data. Benefiting from our rewriting mechanism, new observation-instruction can be obtained in both simulator-free and labor-saving manners to promote generalization. Specifically, we first introduce Object-Enriched Observation Rewriting, where we combine Vision-Language Models (VLMs) and Large Language Models (LLMs) to derive rewritten object-enriched scene descriptions, enabling observation synthesis with diverse objects and spatial layouts via Text-to-Image Generation Models (T2IMs). Then, we propose Observation-Contrast Instruction Rewriting, which generates observation-aligned rewritten instructions by requiring LLMs to reason the difference between original and new observations. We further develop a mixing-then-focusing training strategy with a random observation cropping scheme, effectively enhancing data distribution diversity while suppressing augmentation data noise during training. Experiments on both the discrete environments (R2R, REVERIE, and R4R datasets) and continuous environments (R2R-CE dataset) show the superior performance and impressive generalization ability of our method. Code is available at https://github.com/SaDil13/VLN-RAM.

As large language models (LLMs) like ChatGPT become increasingly integrated into our everyday lives--from customer service and education to creative work and personal productivity--understanding how people interact with these AI systems has become a pressing issue. Despite the widespread use of LLMs, researchers lack standardized tools for systematically studying people's interactions with LLMs. To address this issue, we introduce GPT for Researchers (G4R), or g4r.org, a free website that researchers can use to easily create and integrate a GPT Interface into their studies. At g4r.org, researchers can (1) enable their study participants to interact with GPT (such as ChatGPT), (2) customize GPT Interfaces to guide participants' interactions with GPT (e.g., set constraints on topics or adjust GPT's tone or response style), and (3) capture participants' interactions with GPT by downloading data on messages exchanged between participants and GPT. By facilitating study participants' interactions with GPT and providing detailed data on these interactions, G4R can support research on topics such as consumer interactions with AI agents or LLMs, AI-assisted decision-making, and linguistic patterns in human-AI communication. With this goal in mind, we provide a step-by-step guide to using G4R at g4r.org.

In this digital world, people freely express their emotions using different social media platforms. As a result, modeling and integrating emotion-understanding models are vital for various human-computer interaction tasks such as decision-making, product and customer feedback analysis, political promotions, marketing research, and social media monitoring. As users express different emotions simultaneously in a single instance, annotating emotions in a multilabel setting such as the EthioEmo (Belay et al., 2025) dataset effectively captures this dynamic. Additionally, incorporating intensity, or the degree of emotion, is crucial, as emotions can significantly differ in their expressive strength and impact. This intensity is significant for assessing whether further action is necessary in decision-making processes, especially concerning negative emotions in applications such as healthcare and mental health studies. To enhance the EthioEmo dataset, we include annotations for the intensity of each labeled emotion. Furthermore, we evaluate various state-of-the-art encoder-only Pretrained Language Models (PLMs) and decoder-only Large Language Models (LLMs) to provide comprehensive benchmarking.

Natural language as a medium for human-computer interaction has long been anticipated, has been undergoing a sea-change with the advent of Large Language Models (LLMs) with startling capacities for processing and generating language. Many of us now treat LLMs as modern-day oracles, asking it almost any kind of question. Unlike its Delphic predecessor, consulting an LLM does not have to be a single-turn activity (ask a question, receive an answer, leave); and -- also unlike the Pythia -- it is widely acknowledged that answers from LLMs can be improved with additional context. In this paper, we aim to study when we need multi-turn interactions with LLMs to successfully get a question answered; or conclude that a question is unanswerable. We present a neural symbolic framework that models the interactions between human and LLM agents. Through the proposed framework, we define incompleteness and ambiguity in the questions as properties deducible from the messages exchanged in the interaction, and provide results from benchmark problems, in which the answer-correctness is shown to depend on whether or not questions demonstrate the presence of incompleteness or ambiguity (according to the properties we identify). Our results show multi-turn interactions are usually required for datasets which have a high proportion of incompleteness or ambiguous questions; and that that increasing interaction length has the effect of reducing incompleteness or ambiguity. The results also suggest that our measures of incompleteness and ambiguity can be useful tools for characterising interactions with an LLM on question-answeringproblems

In this paper, we establish a benchmark for evaluating large language models (LLMs) on multi-step geospatial tasks relevant to commercial GIS practitioners. We assess seven leading commercial LLMs (Sonnet 3.5 and 3.7, Haiku 3.5, Gemini 2.0, GPT-4o, GPT-4o mini, and o3-mini) using a simple tool-calling agent equipped with 23 geospatial functions. Our benchmark comprises tasks across four categories of increasing complexity, with both solvable and intentionally unsolvable tasks to test hallucination rejection. We develop an LLM-as-Judge evaluation framework to compare agent solutions against reference implementations. Results show Sonnet 3.5 and GPT-4o achieve the best overall performance, with Claude models excelling on solvable tasks while OpenAI models better identify unsolvable scenarios. We observe significant differences in token usage, with Anthropic models consuming substantially more tokens than competitors. Common errors include misunderstanding geometrical relationships, relying on outdated knowledge, and inefficient data manipulation. The resulting benchmark set, evaluation framework, and data generation pipeline are released as open-source resources, providing one more standardized method for ongoing evaluation of LLMs for GeoAI.

Retrieval-augmented generation (RAG) faces challenges related to factual correctness, source attribution, and response completeness. To address them, we propose a modular pipeline for grounded response generation that operates on information nuggets-minimal, atomic units of relevant information extracted from retrieved documents. The multistage pipeline encompasses nugget detection, clustering, ranking, top cluster summarization, and fluency enhancement. It guarantees grounding in specific facts, facilitates source attribution, and ensures maximum information inclusion within length constraints. Extensive experiments on the TREC RAG'24 dataset evaluated with the AutoNuggetizer framework demonstrate that GINGER achieves state-of-the-art performance on this benchmark.

Large Language Models (LLMs) have shown remarkable proficiency in Machine Reading Comprehension (MRC) tasks; however, their effectiveness for low-resource languages like Vietnamese remains largely unexplored. In this paper, we fine-tune and evaluate two state-of-the-art LLMs: Llama 3 (8B parameters) and Gemma (7B parameters), on ViMMRC, a Vietnamese MRC dataset. By utilizing Quantized Low-Rank Adaptation (QLoRA), we efficiently fine-tune these models and compare their performance against powerful LLM-based baselines. Although our fine-tuned models are smaller than GPT-3 and GPT-3.5, they outperform both traditional BERT-based approaches and these larger models. This demonstrates the effectiveness of our fine-tuning process, showcasing how modern LLMs can surpass the capabilities of older models like BERT while still being suitable for deployment in resource-constrained environments. Through intensive analyses, we explore various aspects of model performance, providing valuable insights into adapting LLMs for low-resource languages like Vietnamese. Our study contributes to the advancement of natural language processing in low-resource languages, and we make our fine-tuned models publicly available at: https://huggingface.co/iaiuet.

In this work, we present the first open leaderboard for evaluating Korean large language models focused on finance. Operated for about eight weeks, the leaderboard evaluated 1,119 submissions on a closed benchmark covering five MCQA categories: finance and accounting, stock price prediction, domestic company analysis, financial markets, and financial agent tasks and one open-ended qa task. Building on insights from these evaluations, we release an open instruction dataset of 80k instances and summarize widely used training strategies observed among top-performing models. Finally, we introduce Won, a fully open and transparent LLM built using these best practices. We hope our contributions help advance the development of better and safer financial LLMs for Korean and other languages.