Large language models achieve exceptional performance on various downstream tasks through supervised fine-tuning. However, the diversity of downstream tasks and practical requirements makes deploying multiple full-parameter fine-tuned models challenging. Current methods that compress the delta weight struggle to achieve ultra-high compression, failing to minimize the deployment overhead. To address the above issue, we propose a novel distribution-driven delta compression framework DeltaDQ, which utilizes Group-wise Dropout and Separate Quantization to achieve ultra-high compression for the delta weight. We have observed that the matrix-computed intermediate results for the delta weight exhibit extremely small variance and min-max range characteristics, referred to as Balanced Intermediate Results. Exploiting this phenomenon, we introduce Group-wise Dropout to perform dropout on the delta weight using an optimal group size. Furthermore, using Separate Quantization, sparse weights are quantized and decomposed to achieve a lower bit. Experimental results show that DeltaDQ achieves 16x compression with improved accuracy compared to baselines for WizardMath and WizardCoder models across different parameter scales. Moreover, DeltaDQ demonstrates the ability for ultra-high compression ratio, achieving 128x compression for the WizardMath-7B model and 512x compression for the WizardMath-70B model.

We test whether replacing crowdsource workers with LLMs to write Natural Language Inference (NLI) hypotheses similarly results in annotation artifacts. We recreate a portion of the Stanford NLI corpus using GPT-4, Llama-2 and Mistral 7b, and train hypothesis-only classifiers to determine whether LLM-elicited hypotheses contain annotation artifacts. On our LLM-elicited NLI datasets, BERT-based hypothesis-only classifiers achieve between 86-96% accuracy, indicating these datasets contain hypothesis-only artifacts. We also find frequent "give-aways" in LLM-generated hypotheses, e.g. the phrase "swimming in a pool" appears in more than 10,000 contradictions generated by GPT-4. Our analysis provides empirical evidence that well-attested biases in NLI can persist in LLM-generated data.

We present \llinstruct: An 8B instruction-tuned model that is designed to generate content for English Language Proficiency Assessments (ELPA) and related applications. Our work involves creating a new dataset of 70K instructions and explanations in the ELPA domain and using these to fine-tune Llama-3 8B models (SFT) of different sizes (e.g., SFT-17K, SFT-50K and SFT-70K). Human evaluations are conducted over unseen instructions to compare these SFT models against SOTA models (e.g., Dolly-2, Mistral, Llama-3 base version, and GPT-3.5). The findings show although all three SFT models perform comparably, the model trained on largest instruction dataset -- SFT-70K - leads to the most valid outputs ready for assessments. However, although the SFT models perform better than larger model, e.g., GPT 3.5 on the aspect of explanations of outputs, many outputs still need human interventions to make them actual ready for real world assessments.

ChatGPT, exploring how structured prompts and active TA support can enhance educational outcomes. Our work also integrates Integrating large language models (LLMs) like ChatGPT is revolutionizing two versions of ChatGPT--ChatGPT-4o and ChatGPT o1--each contributing the field of computer science education. These models uniquely: ChatGPT-4o supports routine educational tasks, offer new possibilities for enriching student learning and supporting while ChatGPT o1 enhances complex problem-solving through improved teaching assistants (TAs) in providing prompt feedback and reasoning, thereby addressing the limitations of using LLMs supplementary learning resources. This research delves into the use independently. of ChatGPT in a data structures and algorithms (DSA) course, particularly when combined with TA supervision. The findings demonstrate that incorporating ChatGPT with structured prompts and RQ1: How does ChatGPT impact students' learning and active TA guidance enhances students' understanding of intricate exam readiness when supervised by TAs? algorithmic concepts, boosts engagement, and elevates academic performance. However, challenges exist in addressing academic RQ2: What challenges arise in utilizing ChatGPT to answer integrity and the limitations of LLMs in tackling complex problems.

Knowledge graph-grounded dialog generation requires retrieving a dialog-relevant subgraph from the given knowledge base graph and integrating it with the dialog history. Previous works typically represent the graph using an external encoder, such as graph neural networks, and retrieve relevant triplets based on the similarity between single-vector representations of triplets and the dialog history. However, these external encoders fail to leverage the rich knowledge of pretrained language models, and the retrieval process is also suboptimal due to the information bottleneck caused by the single-vector abstraction of the dialog history. In this work, we propose Dialog generation with Generative Subgraph Retrieval (DialogGSR), which retrieves relevant knowledge subgraphs by directly generating their token sequences on top of language models. For effective generative subgraph retrieval, we introduce two key methods: (i) structure-aware knowledge graph linearization with self-supervised graph-specific tokens and (ii) graph-constrained decoding utilizing graph structural proximity-based entity informativeness scores for valid and relevant generative retrieval. DialogGSR achieves state-of-the-art performance in knowledge graph-grounded dialog generation, as demonstrated on OpenDialKG and KOMODIS datasets.

They appear extensively across all domains of natural language, from the most sophisticated poetry to seemingly dry academic prose. A significant body of research in the cognitive science of language argues for the existence of conceptual metaphors, the systematic structuring of one domain of experience in the language of another. Conceptual metaphors are not simply rhetorical flourishes but are crucial evidence of the role of analogical reasoning in human cognition. In this paper, we ask whether Large Language Models (LLMs) can accurately identify and explain the presence of such conceptual metaphors in natural language data. Using a novel prompting technique based on metaphor annotation guidelines, we demonstrate that LLMs are a promising tool for large-scale computational research on conceptual metaphors. Further, we show that LLMs are able to apply procedural guidelines designed for human annotators, displaying a surprising depth of linguistic knowledge.

Ptychography is an advanced computational imaging technique in X-ray and electron microscopy. It has been widely adopted across scientific research fields, including physics, chemistry, biology, and materials science, as well as in industrial applications such as semiconductor characterization. In practice, obtaining high-quality ptychographic images requires simultaneous optimization of numerous experimental and algorithmic parameters. Traditionally, parameter selection often relies on trial and error, leading to low-throughput workflows and potential human bias. In this work, we develop the "Ptychographic Experiment and Analysis Robot" (PEAR), a framework that leverages large language models (LLMs) to automate data analysis in ptychography. To ensure high robustness and accuracy, PEAR employs multiple LLM agents for tasks including knowledge retrieval, code generation, parameter recommendation, and image reasoning. Our study demonstrates that PEAR's multi-agent design significantly improves the workflow success rate, even with smaller open-weight models such as LLaMA 3.1 8B. PEAR also supports various automation levels and is designed to work with customized local knowledge bases, ensuring flexibility and adaptability across different research environments.

The current paradigm for safety alignment of large language models (LLMs) follows a one-size-fits-all approach: the model refuses to interact with any content deemed unsafe by the model provider. This approach lacks flexibility in the face of varying social norms across cultures and regions. In addition, users may have diverse safety needs, making a model with static safety standards too restrictive to be useful, as well as too costly to be re-aligned. We propose Controllable Safety Alignment (CoSA), a framework designed to adapt models to diverse safety requirements without re-training. Instead of aligning a fixed model, we align models to follow safety configs -- free-form natural language descriptions of the desired safety behaviors -- that are provided as part of the system prompt. To adjust model safety behavior, authorized users only need to modify such safety configs at inference time. To enable that, we propose CoSAlign, a data-centric method for aligning LLMs to easily adapt to diverse safety configs. Furthermore, we devise a novel controllability evaluation protocol that considers both helpfulness and configured safety, summarizing them into CoSA-Score, and construct CoSApien, a human-authored benchmark that consists of real-world LLM use cases with diverse safety requirements and corresponding evaluation prompts. We show that CoSAlign leads to substantial gains of controllability over strong baselines including in-context alignment. Our framework encourages better representation and adaptation to pluralistic human values in LLMs, and thereby increasing their practicality.

Using Large Language Models (LLMs) to generate database queries is an area of active research. In [4], Sequeda et al. argue that knowledge graphs (KGs) with rich ontologies can enable an LLM to answer queries of enterprise complexity, noting that text-to-SQL benchmarks such as Spider [6] are not tailored to such queries. In addition to query complexity, an equally challenging problem in the enterprise setting is schema complexity, where the ontology itself is large and complex. This paper contributes an approach to using off-the-shelf LLMs and enterprise-scale ontologies to answer natural language questions on large data sets. We address the schema complexity problem by incrementally revealing "just enough" of an ontology that is needed to answer a given question.

Questionnaires are a common method for detecting the personality of Large Language Models (LLMs). However, their reliability is often compromised by two main issues: hallucinations (where LLMs produce inaccurate or irrelevant responses) and the sensitivity of responses to the order of the presented options. To address these issues, we propose combining text mining with questionnaires method. Text mining can extract psychological features from the LLMs' responses without being affected by the order of options. Furthermore, because this method does not rely on specific answers, it reduces the influence of hallucinations. By normalizing the scores from both methods and calculating the root mean square error, our experiment results confirm the effectiveness of this approach. To further investigate the origins of personality traits in LLMs, we conduct experiments on both pre-trained language models (PLMs), such as BERT and GPT, as well as conversational models (ChatLLMs), such as ChatGPT. The results show that LLMs do contain certain personalities, for example, ChatGPT and ChatGLM exhibit the personality traits of 'Conscientiousness'. Additionally, we find that the personalities of LLMs are derived from their pre-trained data. The instruction data used to train ChatLLMs can enhance the generation of data containing personalities and expose their hidden personality. We compare the results with the human average personality score, and we find that the personality of FLAN-T5 in PLMs and ChatGPT in ChatLLMs is more similar to that of a human, with score differences of 0.34 and 0.22, respectively.