This paper explores how non-experts can teach robots desired skills in their environments. We argue that natural language is an intuitive and accessible interface for robot learning. To this end, we investigate two key aspects: (1) how non-experts collect robotic data using natural language supervision and (2) how pre-trained vision-language models learn end-to-end policies directly from this supervision. We propose a data collection framework that collects robot demonstrations based on natural language supervision (e.g., "move forward") and further augments these demonstrations. Next, we introduce a model that learns language-conditioned policies from natural language supervision called CLIP-RT. Our model employs pre-trained CLIP models and learns to predict actions represented in language via contrastive imitation learning. We first train CLIP-RT on large-scale robotic data and then enable it to learn desired skills using data collected from our framework. CLIP-RT shows strong capabilities in acquiring novel manipulation skills, outperforming the state-of-the-art model, OpenVLA (7B parameters), by 17% in average success rates, while using 7x fewer parameters (1B).

The increasing prevalence of AI-generated content alongside human-written text underscores the need for reliable discrimination methods. To address this challenge, we propose a novel framework with textual embeddings from Pre-trained Language Models (PLMs) to distinguish AI-generated and human-authored text. Our approach utilizes Embedding Fusion to integrate semantic information from multiple Language Models, harnessing their complementary strengths to enhance performance. Through extensive evaluation across publicly available diverse datasets, our proposed approach demonstrates strong performance, achieving classification accuracy greater than 96% and a Matthews Correlation Coefficient (MCC) greater than 0.93. This evaluation is conducted on a balanced dataset of texts generated from five well-known Large Language Models (LLMs), highlighting the effectiveness and robustness of our novel methodology.

Large Language Models (LLMs) have achieved state-of-the-art performance across numerous tasks. However, these advancements have predominantly benefited "first-class" languages such as English and Chinese, leaving many other languages underrepresented. This imbalance, while limiting broader applications, generates a natural preference ranking between languages, offering an opportunity to bootstrap the multilingual capabilities of LLM in a self-improving manner. Thus, we propose Language Imbalance Driven Rewarding, where the inherent imbalance between dominant and non-dominant languages within LLMs is leveraged as a reward signal. Iterative DPO training demonstrates that this approach not only enhances LLM performance in non-dominant languages but also improves the dominant language's capacity, thereby yielding an iterative reward signal. Fine-tuning Meta-Llama-3-8B-Instruct over two iterations of this approach results in continuous improvements in multilingual performance across instruction-following and arithmetic reasoning tasks, evidenced by an average improvement of 7.46% win rate on the X-AlpacaEval leaderboard and 13.9% accuracy on the MGSM benchmark. This work serves as an initial exploration, paving the way for multilingual self-improvement of LLMs. Large Language Models (LLMs) have revolutionized the field of Natural Language Processing (NLP) with superior performance across numerous tasks. However, existing studies show that due to the imbalance of pre-training and fine-tuning data across languages, existing LLMs have predominately benefited a few "first-class" languages, particularly English and Chinese, thereby overlooking a wide range of other languages (Qin et al., 2024). Given that LLMs are used worldwide, such language imbalance presents significant risks for users who operate in less dominant languages (Deshpande et al., 2023). To this end, enhancing the multilingual performance of LLMs has gained increasing attention. Previous research predominantly frames this imbalance as an issue to be resolved, often addressing it through multilingual training and cross-lingual alignment.

Evaluations are critical for understanding the capabilities of large language models (LLMs). Fundamentally, evaluations are experiments; but the literature on evaluations has largely ignored the literature from other sciences on experiment analysis and planning. This article shows researchers with some training in statistics how to think about and analyze data from language model evaluations. Conceptualizing evaluation questions as having been drawn from an unseen super-population, we present formulas for analyzing evaluation data, measuring differences between two models, and planning an evaluation experiment. We make a number of specific recommendations for running language model evaluations and reporting experiment results in a way that minimizes statistical noise and maximizes informativeness.

Advances in large language models (LLMs) have spurred research into enhancing their reasoning capabilities, particularly in math-rich STEM documents. While LLMs can generate equations or solve math-related queries, their ability to fully understand and interpret abstract mathematical symbols in long, math-rich documents remains limited. In this paper, we introduce STEM-PoM, a comprehensive benchmark dataset designed to evaluate LLMs' reasoning abilities on math symbols within contextual scientific text. The dataset, sourced from real-world ArXiv documents, contains over 2K math symbols classified as main attributes of variables, constants, operators, and unit descriptors, with additional sub-attributes including scalar/vector/matrix for variables and local/global/discipline-specific labels for both constants and operators. Our extensive experiments show that state-of-the-art LLMs achieve an average of 20-60% accuracy under in-context learning and 50-60% accuracy with fine-tuning, revealing a significant gap in their mathematical reasoning capabilities. STEM-PoM fuels future research of developing advanced Math-AI models that can robustly handle math symbols.

Instruction tuning is a supervised fine-tuning approach that significantly improves the ability of large language models (LLMs) to follow human instructions. We propose SelfCodeAlign, the first fully transparent and permissive pipeline for self-aligning code LLMs without extensive human annotations or distillation. SelfCodeAlign employs the same base model for inference throughout the data generation process. It first extracts diverse coding concepts from high-quality seed snippets to generate new tasks. It then samples multiple responses per task, pairs each with test cases, and validates them in a sandbox environment. Finally, passing examples are selected for instruction tuning. In our primary experiments, we use SelfCodeAlign with CodeQwen1.5-7B to generate a dataset of 74k instruction-response pairs. Finetuning on this dataset leads to a model that achieves a 67.1 pass@1 on HumanEval+, surpassing CodeLlama-70B-Instruct despite being ten times smaller. Across all benchmarks, this finetuned model consistently outperforms the original version trained with OctoPack, the previous state-of-the-art method for instruction tuning without human annotations or distillation. Additionally, we show that SelfCodeAlign is effective across LLMs of various sizes, from 3B to 33B, and that the base models can benefit more from alignment with their own data distribution. We further validate each component's effectiveness in our pipeline, showing that SelfCodeAlign outperforms both direct distillation from GPT-4o and leading GPT-3.5-based distillation methods, such as OSS-Instruct and Evol-Instruct. SelfCodeAlign has also led to the creation of StarCoder2-Instruct, the first fully transparent, permissively licensed, and self-aligned code LLM that achieves state-of-the-art coding performance.

Recent advances in diffusion models have significantly enhanced the quality of image synthesis, yet they have also introduced serious safety concerns, particularly the generation of Not Safe for Work (NSFW) content. Previous research has demonstrated that adversarial prompts can be used to generate NSFW content. However, such adversarial text prompts are often easily detectable by text-based filters, limiting their efficacy. In this paper, we expose a previously overlooked vulnerability: adversarial image attacks targeting Image-to-Image (I2I) diffusion models. We propose AdvI2I, a novel framework that manipulates input images to induce diffusion models to generate NSFW content. By optimizing a generator to craft adversarial images, AdvI2I circumvents existing defense mechanisms, such as Safe Latent Diffusion (SLD), without altering the text prompts. Furthermore, we introduce AdvI2I-Adaptive, an enhanced version that adapts to potential countermeasures and minimizes the resemblance between adversarial images and NSFW concept embeddings, making the attack more resilient against defenses. Through extensive experiments, we demonstrate that both AdvI2I and AdvI2I-Adaptive can effectively bypass current safeguards, highlighting the urgent need for stronger security measures to address the misuse of I2I diffusion models.

The level of detail of the feedback influences its effectiveness [80], and feedback can be given at many levels ranging from targeting how to work on and complete specific tasks to considering personal characteristics and behavior[26, 36, 59]. In teaching and learning programming, automated assessment systems have been a key tool for providing feedback at a scale already for more than a half a century [30, 36, 61]. Researchers have sought to automate step-by-step guidance [78], provide hints during the programming process [55], improve programming error messages [6], and aid in providing textual feedback by grouping similar code submissions together [23, 37, 58]. To support the understanding of how novices construct programs, researchers and educators have been collecting increasing amounts of data from students' programming process [31]. Such data can be collected at multiple granularities, ranging from final course assignment submissions to individual keystrokes from solving the assignments [31]. Programming process data has been, for example, used to play back how students construct their programs step by step or keystroke by keystroke to create a broader understanding of the process [27, 73, 83]. So far, despite shared efforts towards providing timely feedback to students[33], the potential of fine-grained programming process data for feedback purposes is still largely untapped. Large Language Models (LLMs) are a potential tool for realizing the transformation of programming process data into actionable feedback items. Within Computing Education Research, LLMs have broadened the horizon of what computing education researchers and practitioners can achieve[65], calling even for rethinking how computer science and programming is taught [16].

In recent years, large language models (LLMs) have significantly advanced the field of natural language processing (NLP). By fine-tuning LLMs with data from specific scenarios, these foundation models can better adapt to various downstream tasks. However, the fine-tuning process poses privacy leakage risks, particularly in centralized data processing scenarios. To address user privacy concerns, federated learning (FL) has been introduced to mitigate the risks associated with centralized data collection from multiple sources. Nevertheless, the privacy of LLMs themselves is equally critical, as potential malicious attacks challenge their security, an issue that has received limited attention in current research. Consequently, establishing a trusted multi-party model fine-tuning environment is essential. Additionally, the local deployment of large LLMs incurs significant storage costs and high computational demands. To address these challenges, we propose for the first time a federated discrete and transferable prompt tuning, namely FedDTPT, for black-box large language models. In the client optimization phase, we adopt a token-level discrete prompt optimization method that leverages a feedback loop based on prediction accuracy to drive gradient-free prompt optimization through the MLM API. For server optimization, we employ an attention mechanism based on semantic similarity to filter all local prompt tokens, along with an embedding distance elbow detection and DBSCAN clustering strategy to enhance the filtering process. Experimental results demonstrate that, compared to state-of-the-art methods, our approach achieves higher accuracy, reduced communication overhead, and robustness to non-iid data in a black-box setting. Moreover, the optimized prompts are transferable.

In question-answering scenarios, humans can assess whether the available information is sufficient and seek additional information if necessary, rather than providing a forced answer. In contrast, Vision Language Models (VLMs) typically generate direct, one-shot responses without evaluating the sufficiency of the information. To investigate this gap, we identify a critical and challenging task in the Visual Question Answering (VQA) scenario: can VLMs indicate how to adjust an image when the visual information is insufficient to answer a question? This capability is especially valuable for assisting visually impaired individuals who often need guidance to capture images correctly. To evaluate this capability of current VLMs, we introduce a human-labeled dataset as a benchmark for this task. Additionally, we present an automated framework that generates synthetic training data by simulating ``where to know'' scenarios. Our empirical results show significant performance improvements in mainstream VLMs when fine-tuned with this synthetic data. This study demonstrates the potential to narrow the gap between information assessment and acquisition in VLMs, bringing their performance closer to humans.