Direct Preference Optimization (DPO) is widely utilized in the Reinforcement Learning from Human Feedback (RLHF) phase to align Large Language Models (LLMs) with human preferences, thereby enhancing both their harmlessness and efficacy. However, it has been observed that DPO tends to over-optimize for verbosity, which can detrimentally affect both performance and user experience. In this paper, we conduct an in-depth theoretical analysis of DPO's optimization objective and reveal a strong correlation between its implicit reward and data length. To address this issue, we propose a length-desensitization improvement method for DPO, termed LD-DPO. The proposed method aims to desensitize DPO to data length by decoupling explicit length preference, which is relatively insignificant, from the other implicit preferences, thereby enabling more effective learning of the intrinsic preferences. We utilized two settings (Base and Instruct) of Llama2-13B, Llama3-8B, and Qwen2-7B for experimental validation on various benchmarks including MT-Bench and AlpacaEval 2. The experimental results indicate that LD-DPO consistently outperforms DPO and other baseline methods, achieving more concise responses with a 10-40% reduction in length compared to DPO. We conducted in-depth experimental analyses to demonstrate that LD-DPO can indeed achieve length desensitization and align the model more closely with human-real preferences. "Brevity is the Soul of Wit." Human preference alignment is crucial to enable large language models (LLMs) to be helpful, honest, and harmless. Among the various methods to achieve effective alignment (Dai et al., 2024; Yuan et al., 2024a), Directed Preference Optimization (DPO) has emerged as a promising technique (Rafailov et al., 2024), giving rise to numerous derivative algorithms (Hong et al., 2024; Chen et al., 2024b; Ethayarajh et al., 2024). DPO eliminates the reliance on online Reward Models (RMs) by reparameterizing the reward function in Reinforcement Learning from Human Feedback (RLHF), thereby implementing a simple and stable offline preference learning paradigm. Among the dimensions of human language preferences, detailedness is one of the most straightforward categories that current alignment algorithms can effortlessly capture, as longer texts tend to be richer in content. However, it has been demonstrated that DPO is susceptible to an over-optimization issue in this particular preference dimension (Xu et al., 2024). As shown in Fig.1, this overemphasis results in models that produce excessively verbose responses, which can compromise their instruction-following and reasoning capabilities (Ding et al., 2023; Yuan et al., 2024b).

In the real world, where information is abundant and diverse across different modalities, understanding and utilizing various data types to improve retrieval systems is a key focus of research. Multimodal composite retrieval integrates diverse modalities such as text, image and audio, etc. to provide more accurate, personalized, and contextually relevant results. To facilitate a deeper understanding of this promising direction, this survey explores multimodal composite editing and retrieval in depth, covering image-text composite editing, image-text composite retrieval, and other multimodal composite retrieval. In this survey, we systematically organize the application scenarios, methods, benchmarks, experiments, and future directions. Multimodal learning is a hot topic in large model era, and have also witnessed some surveys in multimodal learning and vision-language models with transformers published in the PAMI journal. To the best of our knowledge, this survey is the first comprehensive review of the literature on multimodal composite retrieval, which is a timely complement of multimodal fusion to existing reviews. To help readers' quickly track this field, we build the project page for this survey, which can be found at https://github.com/fuxianghuang1/Multimodal-Composite-Editing-and-Retrieval.

Large Language Models (LLM) often needs to be Continual Pre-Trained (CPT) to obtain the unfamiliar language skill or adapt into new domains. The huge training cost of CPT often asks for cautious choice of key hyper-parameters such as the mixture ratio of extra language or domain corpus. However, there is no systematic study which bridge the gap between the optimal mixture ratio and the actual model performance, and the gap between experimental scaling law and the actual deployment in the full model size. In this paper, we perform CPT on Llama-3 8B and 70B to enhance its Chinese ability. We study the optimal correlation between the Additional Language Mixture Ratio (ALMR) and the Learning Rate (LR) on the 8B size which directly indicate the optimal experimental set up. By thorough choice of hyper-parameter, and subsequent fine-tuning, the model capability is improved not only on the Chinese-related benchmark, but also some specific domains including math, coding and emotional intelligence. We deploy the final 70B version of LLM on an real-life chat system which obtain satisfying performance.

Sequence recommendation (SeqRec) aims to predict the next item a user will interact with by understanding user intentions and leveraging collaborative filtering information. Large language models (LLMs) have shown great promise in recommendation tasks through prompt-based, fixed reflection libraries, and fine-tuning techniques. However, these methods face challenges, including lack of supervision, inability to optimize reflection sources, inflexibility to diverse user needs, and high computational costs. Despite promising results, current studies primarily focus on reflections of users' explicit preferences (e.g., item titles) while neglecting implicit preferences (e.g., brands) and collaborative filtering information. This oversight hinders the capture of preference shifts and dynamic user behaviors. Additionally, existing approaches lack mechanisms for reflection evaluation and iteration, often leading to suboptimal recommendations. To address these issues, we propose the Mixture of REflectors (MoRE) framework, designed to model and learn dynamic user preferences in SeqRec. Specifically, MoRE introduces three reflectors for generating LLM-based reflections on explicit preferences, implicit preferences, and collaborative signals. Each reflector incorporates a self-improving strategy, termed refining-and-iteration, to evaluate and iteratively update reflections. Furthermore, a meta-reflector employs a contextual bandit algorithm to select the most suitable expert and corresponding reflections for each user's recommendation, effectively capturing dynamic preferences. Extensive experiments on three real-world datasets demonstrate that MoRE consistently outperforms state-of-the-art methods, requiring less training time and GPU memory compared to other LLM-based approaches in SeqRec.

Given the profound impact of narratives across various societal levels, from personal identities to international politics, it is crucial to understand their distribution and development over time. This is particularly important in online spaces. On the Web, narratives can spread rapidly and intensify societal divides and conflicts. While many qualitative approaches exist, quantifying narratives remains a significant challenge. Computational narrative analysis lacks frameworks that are both comprehensive and generalizable. To address this gap, we introduce a numerical narrative representation grounded in structuralist linguistic theory. Chiefly, Greimas' Actantial Model represents a narrative through a constellation of six functional character roles. These so-called actants are genre-agnostic, making the model highly generalizable. We extract the actants using an open-source LLM and integrate them into a Narrative-Structured Text Embedding that captures both the semantics and narrative structure of a text. We demonstrate the analytical insights of the method on the example of 5000 full-text news articles from Al Jazeera and The Washington Post on the Israel-Palestine conflict. Our method successfully distinguishes articles that cover the same topics but differ in narrative structure.

Much of the cost and effort required during the software testing process is invested in performing test maintenance - the addition, removal, or modification of test cases to keep the test suite in sync with the system-under-test or to otherwise improve its quality. Tool support could reduce the cost - and improve the quality - of test maintenance by automating aspects of the process or by providing guidance and support to developers. In this study, we explore the capabilities and applications of large language models (LLMs) - complex machine learning models adapted to textual analysis - to support test maintenance. We conducted a case study at Ericsson AB where we explored the triggers that indicate the need for test maintenance, the actions that LLMs can take, and the considerations that must be made when deploying LLMs in an industrial setting. We also proposed and demonstrated implementations of two multi-agent architectures that can predict which test cases require maintenance following a change to the source code. Collectively, these contributions advance our theoretical and practical understanding of how LLMs can be deployed to benefit industrial test maintenance processes.

Many algorithms for aligning LLMs with human preferences assume that human preferences are binary and deterministic. However, it is reasonable to think that they can vary with different individuals, and thus should be distributional to reflect the fine-grained relationship between the responses. In this work, we introduce the distributional soft preference labels and improve Direct Preference Optimization (DPO) with a weighted geometric average of the LLM output likelihood in the loss function. In doing so, the scale of learning loss is adjusted based on the soft labels, and the loss with equally preferred responses would be close to zero. This simple modification can be easily applied to any DPO family and helps the models escape from the over-optimization and objective mismatch prior works suffer from. In our experiments, we simulate the soft preference labels with AI feedback from LLMs and demonstrate that geometric averaging consistently improves performance on standard benchmarks for alignment research. In particular, we observe more preferable responses than binary labels and significant improvements with data where modestly-confident labels are in the majority.

The remarkable multimodal capabilities and interactive experience of GPT-4o underscore their necessity in practical applications, yet open-source models rarely excel in both areas. In this paper, we introduce VITA, the first-ever open-source Multimodal Large Language Model (MLLM) adept at simultaneous processing and analysis of Video, Image, Text, and Audio modalities, and meanwhile has an advanced multimodal interactive experience. Starting from Mixtral 8x7B as a language foundation, we expand its Chinese vocabulary followed by bilingual instruction tuning. We further endow the language model with visual and audio capabilities through two-stage multi-task learning of multimodal alignment and instruction tuning. VITA demonstrates robust foundational capabilities of multilingual, vision, and audio understanding, as evidenced by its strong performance across a range of both unimodal and multimodal benchmarks. Beyond foundational capabilities, we have made considerable progress in enhancing the natural multimodal human-computer interaction experience. VITA is the first step for the open-source community to explore the seamless integration of multimodal understanding and interaction. While there is still lots of work to be done on VITA to get close to close-source counterparts, we hope that its role as a pioneer can serve as a cornerstone for subsequent research. Project Page: https://vita-home.github.io.

In many industrial settings, users wish to ask questions whose answers may be found in structured data sources such as a spreadsheets, databases, APIs, or combinations thereof. Often, the user doesn't know how to identify or access the right data source. This problem is compounded even further if multiple (and potentially siloed) data sources must be assembled to derive the answer. Recently, various Text-to-SQL applications that leverage Large Language Models (LLMs) have addressed some of these problems by enabling users to ask questions in natural language. However, these applications remain impractical in realistic industrial settings because they fail to cope with the data source heterogeneity that typifies such environments. In this paper, we address heterogeneity by introducing the siwarex platform, which enables seamless natural language access to both databases and APIs. To demonstrate the effectiveness of siwarex, we extend the popular Spider dataset and benchmark by replacing some of its tables by data retrieval APIs. We find that siwarex does a good job of coping with data source heterogeneity. Our modified Spider benchmark will soon be available to the research community

This paper introduces text-to-shape-display, a novel approach to generating dynamic shape changes in pin-based shape displays through natural language commands. By leveraging large language models (LLMs) and AI-chaining, our approach allows users to author shape-changing behaviors on demand through text prompts without programming. We describe the foundational aspects necessary for such a system, including the identification of key generative elements (primitive, animation, and interaction) and design requirements to enhance user interaction, based on formative exploration and iterative design processes. Based on these insights, we develop SHAPE-IT, an LLM-based authoring tool for a 24 x 24 shape display, which translates the user's textual command into executable code and allows for quick exploration through a web-based control interface. We evaluate the effectiveness of SHAPE-IT in two ways: 1) performance evaluation and 2) user evaluation (N= 10). The study conclusions highlight the ability to facilitate rapid ideation of a wide range of shape-changing behaviors with AI. However, the findings also expose accuracy-related challenges and limitations, prompting further exploration into refining the framework for leveraging AI to better suit the unique requirements of shape-changing systems.