Large Language Models (LLMs) have significantly advanced various fields, particularly coding, mathematical reasoning, and logical problem solving. However, a critical question remains: Do these mathematical reasoning abilities persist when LLMs are presented with culturally adapted math problems? Specifically, how do LLMs perform when faced with math problems embedded in cultural contexts that have no significant representation in main stream web-scale AI training data? To explore this, we generated six synthetic cultural datasets from GSM8K, a widely used benchmark for assessing LLMs' mathematical reasoning skills. While preserving the mathematical logic and numerical values of the original GSM8K test set, we modify cultural elements such as personal names, food items, place names, etc. These culturally adapted datasets provide a more reliable framework for evaluating LLMs' mathematical reasoning under shifting cultural contexts. Our findings reveal that LLMs struggle with math problems when cultural references change, even though the underlying mathematical structure remains constant. Smaller models exhibit greater performance drops compared to larger models. Interestingly, our results also suggest that cultural familiarity can enhance mathematical reasoning. Even models with no explicit mathematical training but exposure to relevant cultural contexts sometimes outperform larger, mathematically proficient models on culturally embedded math problems. This study highlights the impact of cultural context on the mathematical reasoning abilities of LLMs, underscoring the need for more diverse and representative training data to improve robustness in real-world applications. The benchmark data sets and script for reproducing the results are available at https://github.com/akarim23131/Lost_in_Cultural_Translation

To mitigate misuse, text-to-image~(T2I) models commonly incorporate safety filters to prevent the generation of sensitive images. Unfortunately, recent jailbreaking attack methods use LLMs to generate adversarial prompts that effectively bypass safety filters while generating sensitive images, revealing the safety vulnerabilities within the T2I model. However, existing LLM-based attack methods lack explicit guidance, relying on substantial queries to achieve a successful attack, which limits their practicality in real-world scenarios. In this work, we introduce \textbf{MJA}, a \textbf{m}etaphor-based \textbf{j}ailbreaking \textbf{a}ttack method inspired by the Taboo game, aiming to balance the attack effectiveness and query efficiency by generating metaphor-based adversarial prompts. Specifically, MJA consists of two modules: an LLM-based multi-agent generation module~(MLAG) and an adversarial prompt optimization module~(APO). MLAG decomposes the generation of metaphor-based adversarial prompts into three subtasks: metaphor retrieval, context matching, and adversarial prompt generation. Subsequently, MLAG coordinates three LLM-based agents to generate diverse adversarial prompts by exploring various metaphors and contexts. To enhance the attack efficiency, APO first trains a surrogate model to predict the attack results of adversarial prompts and then designs an acquisition strategy to adaptively identify optimal adversarial prompts. Experiments demonstrate that MJA achieves better attack effectiveness while requiring fewer queries compared to baseline methods. Moreover, our adversarial prompts exhibit strong transferability across various open-source and commercial T2I models. \textcolor{red}{This paper includes model-generated content that may contain offensive or distressing material.}

Large Language Models (LLMs) are computationally intensive, particularly during inference. Neuron-adaptive techniques, which selectively activate neurons in Multi-Layer Perceptron (MLP) layers, offer some speedups but suffer from limitations in modern Transformers. These include reliance on sparse activations, incompatibility with attention layers, and the use of costly neuron masking techniques. To address these issues, we propose the Adaptive Rank Allocation framework and introduce the Rank and Neuron Allocator (RaNA) adapter. RaNA adapters leverage rank adapters, which operate on linear layers by applying both low-rank matrix decompositions and adaptive masking to efficiently allocate compute without depending on activation sparsity. This enables RaNA to be generally applied to MLPs and linear components of attention modules, while eliminating the need for expensive maskers found in neuron-adaptive methods. Notably, when compared to neuron adapters, RaNA improves perplexity by up to 7 points and increases accuracy by up to 8 percentage-points when reducing FLOPs by 44% in state-of-the-art Transformer architectures. As Large Language Models (LLMs) have grown in popularity and size, they have begun consuming a non-trivial amount of compute and time for training and inference (Kim et al. (2023), Pope et al. (2022)). Adaptive compute methods seek to speed up the inference stage of Transformers (Vaswani et al. (2023)), the de facto LLM architecture, by identifying and avoiding redundant computations to save I/O and floating-point operations (FLOPs).

The rapid growth of AI-generated content (AIGC) services has created an urgent need for effective prompt pricing strategies, yet current approaches overlook users' strategic two-step decision-making process in selecting and utilizing generative AI models. This oversight creates two key technical challenges: quantifying the relationship between user prompt capabilities and generation outcomes, and optimizing platform payoff while accounting for heterogeneous user behaviors. We address these challenges by introducing prompt ambiguity, a theoretical framework that captures users' varying abilities in prompt engineering, and developing an Optimal Prompt Pricing (OPP) algorithm. Our analysis reveals a counterintuitive insight: users with higher prompt ambiguity (i.e., lower capability) exhibit non-monotonic prompt usage patterns, first increasing then decreasing with ambiguity levels, reflecting complex changes in marginal utility. Experimental evaluation using a character-level GPT-like model demonstrates that our OPP algorithm achieves up to 31.72% improvement in platform payoff compared to existing pricing mechanisms, validating the importance of user-centric prompt pricing in AIGC services.

Spatial-temporal sequence forecasting (STSF) is a long-standing research problem with widespread real-world applications. Neural architecture search (NAS), which automates the neural network design, has been shown effective in tackling the STSF problem. However, the existing NAS methods for STSF focus on generating architectures in a time-consuming data-driven fashion, which heavily limits their ability to use background knowledge and explore the complicated search trajectory. Large language models (LLMs) have shown remarkable ability in decision-making with comprehensive internal world knowledge, but how it could benefit NAS for STSF remains unexplored. In this paper, we propose a novel NAS method for STSF based on LLM. Instead of directly generate architectures with LLM, We inspire the LLM's capability with a multi-level enhancement mechanism. Specifically, on the step-level, we decompose the generation task into decision steps with powerful prompt engineering and inspire LLM to serve as instructor for architecture search based on its internal knowledge. On the instance-level, we utilize a one-step tuning framework to quickly evaluate the architecture instance and a memory bank to cumulate knowledge to improve LLM's search ability. On the task-level, we propose a two-stage architecture search, balancing the exploration stage and optimization stage, to reduce the possibility of being trapped in local optima. Extensive experimental results demonstrate that our method can achieve competitive effectiveness with superior efficiency against existing NAS methods for STSF.

Large Language Models (LLMs) have demonstrated impressive capabilities across various domains, but their effectiveness in financial decision making, particularly in fund investment, remains inadequately evaluated. Current benchmarks primarily assess LLMs understanding of financial documents rather than their ability to manage assets or analyze trading opportunities in dynamic market conditions. A critical limitation in existing evaluation methodologies is the backtesting approach, which suffers from information leakage when LLMs are evaluated on historical data they may have encountered during pretraining. This paper introduces DeepFund, a comprehensive platform for evaluating LLM based trading strategies in a simulated live environment. Our approach implements a multi agent framework where LLMs serve as both analysts and managers, creating a realistic simulation of investment decision making. The platform employs a forward testing methodology that mitigates information leakage by evaluating models on market data released after their training cutoff dates. We provide a web interface that visualizes model performance across different market conditions and investment parameters, enabling detailed comparative analysis. Through DeepFund, we aim to provide a more accurate and fair assessment of LLMs capabilities in fund investment, offering insights into their potential real world applications in financial markets.

Multi-task prompt tuning utilizes multiple high-resource source tasks to improve performance on low-source target tasks. Existing approaches transfer the soft prompt trained by combining all source tasks or a single ``high-similar'' source task one-time-only. However, we find that the optimal transfer performance often comes from a combination of source tasks, which is neither one nor all. Further, we find that the similarity between source and target tasks also changes dynamically during fine-tuning after transfering, making similarity calculation in the initiation stage inadequate. To address these issues, we propose a method called Dynamic Task Vector Grouping (DTVG), whose core ideas contain (1) measuring the task similarity with task vectors instead of soft prompt, (2) grouping the optimal source task combination based on two metrics: {\it target similarity} and {\it knowledge consistency}; (3) dynamically updating the combination in each iteration step. Extensive experiments on the 26 NLP datasets under different settings demonstrate that DTVG effectively groups similar source tasks while reducing negative transfer, achieving the start-of-art performance.

Reid Hoffman is a prominent Silicon Valley billionaire entrepreneur and investor known for co-founding the professional social networking site LinkedIn, now owned by Microsoft. The longtime Democrat donor threw his support behind Kamala Harris in the race for the White House. Hoffman spoke to the Observer about technology in the new political milieu and his new book about our future with artificial intelligence, Superagency. The book, while not ignoring the problems that AI might cause, argues that the technology is poised to give us cognitive superpowers that will increase our individual and collective human agency, creating a state of widespread empowerment for society. You have a vested interest in being positive about AI, including a company focused on conversational AI for business, Inflection AI.

Higher use of chatbots like ChatGPT may correspond with increased loneliness and less time spent socializing with other people, according to new research from OpenAI in partnership with the Massachusetts Institute of Technology. Those who spent more time typing or speaking with ChatGPT each day tended to report higher levels of emotional dependence on, and problematic use of, the chatbot, as well as heightened levels of loneliness, according to research released Friday. The findings were part of a pair of studies conducted by researchers at the two organizations and have not been peer reviewed. The launch of ChatGPT in late 2022 helped kick off a frenzy for generative artificial intelligence. Since then, people have used chatbots for everything from coding to ersatz therapy sessions.

Language agents powered by large language models (LLMs) face significant deployment challenges in resource-constrained environments, particularly for specialized domains and less-common languages. This paper presents Tox-chat, a Korean chemical toxicity information agent devised within these limitations. We propose two key innovations: a context-efficient architecture that reduces token consumption through hierarchical section search, and a scenario-based dialogue generation methodology that effectively distills tool-using capabilities from larger models. Experimental evaluations demonstrate that our fine-tuned 8B parameter model substantially outperforms both untuned models and baseline approaches, in terms of DB faithfulness and preference. Our work offers valuable insights for researchers developing domain-specific language agents under practical constraints.