Large language models (LLMs) have made significant progress in machine translation (MT). However, their potential in domain-specific MT remains under-explored. Current LLM-based MT systems still face several challenges. First, for LLMs with in-context learning, their effectiveness is highly sensitive to input translation examples, and processing them can increase inference costs. They often require extra post-processing due to over-generation. Second, LLMs with fine-tuning on domain-specific data often require high training costs for domain adaptation, and may weaken the zero-shot MT capabilities of LLMs due to over-specialization. The aforementioned methods can struggle to translate rare words in domain transfer scenarios. To address these challenges, this paper proposes a prompt-oriented fine-tuning method, denoted as LlamaIT, to effectively and efficiently fine-tune a general-purpose LLM for domain-specific MT tasks. First, we construct a task-specific mix-domain dataset, which is then used to fine-tune the LLM with LoRA. This can eliminate the need for input translation examples, post-processing, or over-specialization. By zero-shot prompting with instructions, we adapt the MT tasks to the target domain at inference time. To further elicit the MT capability for rare words, we construct new prompts by incorporating domain-specific bilingual vocabulary. We also conduct extensive experiments on both publicly available and self-constructed datasets. The results show that our LlamaIT can significantly enhance the domain-specific MT capabilities of the LLM, meanwhile preserving its zero-shot MT capabilities.

In online learning platforms, particularly in rapidly growing computer programming courses, addressing the thousands of students' learning queries requires considerable human cost. The creation of intelligent assistant large language models (LLMs) tailored for programming education necessitates distinct data support. However, in real application scenarios, the data resources for training such LLMs are relatively scarce. Therefore, to address the data scarcity in intelligent educational systems for programming, this paper proposes a new Chinese question-and-answer dataset for Python learners. To ensure the authenticity and reliability of the sources of the questions, we collected questions from actual student questions and categorized them according to various dimensions such as the type of questions and the type of learners. This annotation principle is designed to enhance the effectiveness and quality of online programming education, providing a solid data foundation for developing the programming teaching assists (TA). Furthermore, we conducted comprehensive evaluations of various LLMs proficient in processing and generating Chinese content, highlighting the potential limitations of general LLMs as intelligent teaching assistants in computer programming courses.

Tokenization, the division of input text into input tokens, is an often overlooked aspect of the large language model (LLM) pipeline and could be the source of useful or harmful inductive biases. Historically, LLMs have relied on byte pair encoding, without care to specific input domains. With the increased use of LLMs for reasoning, various number-specific tokenization schemes have been adopted, with popular models like LLaMa and PaLM opting for single-digit tokenization while GPT-3.5 and GPT-4 have separate tokens for each 1-, 2-, and 3-digit numbers. In this work, we study the effect this choice has on numerical reasoning through the use of arithmetic tasks. We consider left-to-right and right-to-left tokenization for GPT-3.5 and -4, finding that right-to-left tokenization (enforced by comma separating numbers at inference time) leads to largely improved performance. Furthermore, we find that model errors when using standard left-to-right tokenization follow stereotyped error patterns, suggesting that model computations are systematic rather than approximate. We show that the model is able to convert between tokenizations easily, thus allowing chain-of-thought-inspired approaches to recover performance on left-to-right tokenized inputs. We also find the gap between tokenization directions decreases when models are scaled, possibly indicating that larger models are better able to override this tokenization-dependent inductive bias. In summary, our work performs the first study of how number tokenization choices lead to differences in model performance on arithmetic tasks, accompanied by a thorough analysis of error patterns. We hope this work inspires practitioners to more carefully ablate number tokenization-related choices when working towards general models of numerical reasoning.

The versatility of large language models (LLMs) led to the creation of diverse benchmarks that thoroughly test a variety of language models' abilities. These benchmarks consist of tens of thousands of examples making evaluation of LLMs very expensive. In this paper, we investigate strategies to reduce the number of evaluations needed to assess the performance of an LLM on several key benchmarks. For example, we show that to accurately estimate the performance of an LLM on MMLU, a popular multiple-choice QA benchmark consisting of 14K examples, it is sufficient to evaluate this LLM on 100 curated examples. We release evaluation tools and tiny versions of popular benchmarks: Open LLM Leaderboard, MMLU, HELM, and AlpacaEval 2.0. Our empirical analysis demonstrates that these tools and tiny benchmarks are sufficient to reliably and efficiently reproduce the original evaluation results.

The rise of generative AI tools like ChatGPT has increased the potential for a wide range of attackers to target elections around the world in 2024, according to a new report by cybersecurity giant CrowdStrike. Both state-linked hackers and allied so-called "hacktivists" are increasingly experimenting with ChatGPT and other AI tools, enabling a wider range of actors to carry out cyberattacks and scams, according to the company's annual global threats report. This includes hackers linked to Russia, China, North Korea, and Iran, who have been testing new ways to use these technologies against the U.S., Israel, and European countries. With half the world's population set to vote in 2024, the use of generative AI to target elections could be a "huge factor," says Adam Meyers, head of counter-adversary operations at CrowdStrike. So far, CrowdStrike analysts have been able to detect the use of these models through comments in the scripts that would have been placed there by a tool like ChatGPT.

But Google's Gemini has been accused of being racist towards white people. The tool uses artificial intelligence to create images from prompts within seconds. But users claim the AI bot refuses to create images of Caucasian people, after testing it with requests for Popes, Vikings, and country music fans. 'New game: Try to get Google Gemini to make an image of a Caucasian male. I have not been successful so far,' one user wrote on X (formerly Twitter).

Multimodal reasoning stands as a pivotal capability for large vision-language models (LVLMs). The integration with Domain-Specific Languages (DSL), offering precise visual representations, equips these models with the opportunity to execute more accurate reasoning in complex and professional domains. However, the vanilla Chain-of-Thought (CoT) prompting method faces challenges in effectively leveraging the unique strengths of visual and DSL representations, primarily due to their differing reasoning mechanisms. Additionally, it often falls short in addressing critical steps in multi-step reasoning tasks. To mitigate these challenges, we introduce the \underline{B}i-Modal \underline{B}ehavioral \underline{A}lignment (BBA) prompting method, designed to maximize the potential of DSL in augmenting complex multi-modal reasoning tasks. This method initiates by guiding LVLMs to create separate reasoning chains for visual and DSL representations. Subsequently, it aligns these chains by addressing any inconsistencies, thus achieving a cohesive integration of behaviors from different modalities. Our experiments demonstrate that BBA substantially improves the performance of GPT-4V(ision) on geometry problem solving ($28.34\% \to 34.22\%$), chess positional advantage prediction ($42.08\% \to 46.99\%$) and molecular property prediction ($77.47\% \to 83.52\%$).

In the last years' digitalization process, the creation and management of documents in various domains, particularly in Public Administration (PA), have become increasingly complex and diverse. This complexity arises from the need to handle a wide range of document types, often characterized by semi-structured forms. Semi-structured documents present a fixed set of data without a fixed format. As a consequence, a template-based solution cannot be used, as understanding a document requires the extraction of the data structure. The recent introduction of Large Language Models (LLMs) has enabled the creation of customized text output satisfying user requests. In this work, we propose a novel approach that combines the LLMs with prompt engineering and multi-agent systems for generating new documents compliant with a desired structure. The main contribution of this work concerns replacing the commonly used manual prompting with a task description generated by semantic retrieval from an LLM. The potential of this approach is demonstrated through a series of experiments and case studies, showcasing its effectiveness in real-world PA scenarios.

Recommendation systems are widely used in e-commerce websites and online platforms to address information overload. However, existing systems primarily rely on historical data and user feedback, making it difficult to capture user intent transitions. Recently, Knowledge Base (KB)-based models are proposed to incorporate expert knowledge, but it struggle to adapt to new items and the evolving e-commerce environment. To address these challenges, we propose a novel Large Language Model based Complementary Knowledge Enhanced Recommendation System (LLM-KERec). It introduces an entity extractor that extracts unified concept terms from item and user information. To provide cost-effective and reliable prior knowledge, entity pairs are generated based on entity popularity and specific strategies. The large language model determines complementary relationships in each entity pair, constructing a complementary knowledge graph. Furthermore, a new complementary recall module and an Entity-Entity-Item (E-E-I) weight decision model refine the scoring of the ranking model using real complementary exposure-click samples. Extensive experiments conducted on three industry datasets demonstrate the significant performance improvement of our model compared to existing approaches. Additionally, detailed analysis shows that LLM-KERec enhances users' enthusiasm for consumption by recommending complementary items. In summary, LLM-KERec addresses the limitations of traditional recommendation systems by incorporating complementary knowledge and utilizing a large language model to capture user intent transitions, adapt to new items, and enhance recommendation efficiency in the evolving e-commerce landscape.

Large language models (LLMs) such as ChatGPT have gained considerable interest across diverse research communities. Their notable ability for text completion and generation has inaugurated a novel paradigm for language-interfaced problem solving. However, the potential and efficacy of these models in bioinformatics remain incompletely explored. In this work, we study the performance LLMs on a wide spectrum of crucial bioinformatics tasks. These tasks include the identification of potential coding regions, extraction of named entities for genes and proteins, detection of antimicrobial and anti-cancer peptides, molecular optimization, and resolution of educational bioinformatics problems. Our findings indicate that, given appropriate prompts, LLMs like GPT variants can successfully handle most of these tasks. In addition, we provide a thorough analysis of their limitations in the context of complicated bioinformatics tasks. In conclusion, we believe that this work can provide new perspectives and motivate future research in the field of LLMs applications, AI for Science and bioinformatics.