Abstract--Recent advancements in Multimodal Large Language Models (MLLMs) underscore the significance of scalable models and data to boost performance, yet this often incurs substantial computational costs. Although the Mixture of Experts (MoE) architecture has been employed to efficiently scale large language and image-text models, these efforts typically involve fewer experts and limited modalities. To address this, our work presents the pioneering attempt to develop a unified MLLM with the MoE architecture, named Uni-MoE that can handle a wide array of modalities. Specifically, it features modality-specific encoders with connectors for a unified multimodal representation. We also implement a sparse MoE architecture within the LLMs to enable efficient training and inference through modality-level data parallelism and expert-level model parallelism. To enhance the multi-expert collaboration and generalization, we present a progressive training strategy: 1) Cross-modality alignment using various connectors with different cross-modality data, 2) Training modality-specific experts with cross-modality instruction data to activate experts' preferences, and 3) Tuning the Uni-MoE framework utilizing Low-Rank Adaptation (LoRA) on mixed multimodal instruction data. We evaluate the instruction-tuned Uni-MoE on a comprehensive set of multimodal datasets. The extensive experimental results demonstrate Uni-MoE's principal advantage of significantly reducing performance bias in handling mixed multimodal datasets, alongside improved multi-expert collaboration and generalization. Additionally, there is a growing trend [6], [7], [8], [9] toward building a unified MLLM that could comprehend more modalities such as video, audio, and speech, moving beyond the traditional imagetext paradigm. To catch up with superior closed-source MLLMs like GPT-4V [10] and Gemini [11], the main efforts of open-source community contain enlarging model sizes [12], as seen with the expansion of vision foundation models to 6 billion parameters [12] and the integration with 70B Large Language models (LLMs) [13], [14], and enhancing instruction tuning with diverse multimodal datasets [3], [15], [16]. These developments underscore the increasing ability of MLLMs to process and reason across multiple modalities, showing the importance of both model scalability and the expansion of multimodal instructional data.

Recent advancements in large language models (LLMs) have greatly improved code generation, specifically at the function level. For instance, GPT-4 has achieved an 88.4% pass rate on HumanEval. However, this draws into question the adequacy of existing benchmarks in thoroughly assessing function-level code generation capabilities. Our study analyzed two common benchmarks, HumanEval and MBPP, and found that these might not thoroughly evaluate LLMs' code generation capacities due to limitations in quality, difficulty, and granularity. To resolve this, we introduce the Mostly Hard Python Problems (MHPP) dataset, consisting of 140 unique human-curated problems. By focusing on the combination of natural language and code reasoning, MHPP gauges LLMs' abilities to comprehend specifications and restrictions, engage in multi-step reasoning, and apply coding knowledge effectively. Initial evaluations of 22 LLMs using MHPP showed many high-performing models on HumanEval failed to achieve similar success on MHPP. Moreover, MHPP highlighted various previously undiscovered limitations within various LLMs, leading us to believe that it could pave the way for a better understanding of LLMs' capabilities and limitations.

Conceptualization, or viewing entities and situations as instances of abstract concepts in mind and making inferences based on that, is a vital component in human intelligence for commonsense reasoning. Despite recent progress in artificial intelligence to acquire and model commonsense attributed to neural language models and commonsense knowledge graphs (CKGs), conceptualization is yet to be introduced thoroughly, making current approaches ineffective to cover knowledge about countless diverse entities and situations in the real world. To address the problem, we thoroughly study the role of conceptualization in commonsense reasoning, and formulate a framework to replicate human conceptual induction by acquiring abstract knowledge about events regarding abstract concepts, as well as higher-level triples or inferences upon them. We then apply the framework to ATOMIC, a large-scale human-annotated CKG, aided by the taxonomy Probase. We annotate a dataset on the validity of contextualized conceptualizations from ATOMIC on both event and triple levels, develop a series of heuristic rules based on linguistic features, and train a set of neural models to generate and verify abstract knowledge. Based on these components, a pipeline to acquire abstract knowledge is built. A large abstract CKG upon ATOMIC is then induced, ready to be instantiated to infer about unseen entities or situations. Finally, we empirically show the benefits of augmenting CKGs with abstract knowledge in downstream tasks like commonsense inference and zero-shot commonsense QA.

The shortage of clinical workforce presents significant challenges in mental healthcare, limiting access to formal diagnostics and services. We aim to tackle this shortage by integrating a customized large language model (LLM) into the workflow, thus promoting equity in mental healthcare for the general population. Although LLMs have showcased their capability in clinical decision-making, their adaptation to severe conditions like Post-traumatic Stress Disorder (PTSD) remains largely unexplored. Therefore, we collect 411 clinician-administered diagnostic interviews and devise a novel approach to obtain high-quality data. Moreover, we build a comprehensive framework to automate PTSD diagnostic assessments based on interview contents by leveraging two state-of-the-art LLMs, GPT-4 and Llama-2, with potential for broader clinical diagnoses. Our results illustrate strong promise for LLMs, tested on our dataset, to aid clinicians in diagnostic validation. To the best of our knowledge, this is the first AI system that fully automates assessments for mental illness based on clinician-administered interviews.

Layout design generation has recently gained significant attention due to its potential applications in various fields, including UI, graphic, and floor plan design. However, existing models face two main challenges that limits their adoption in practice. Firstly, the limited expressiveness of individual condition types used in previous works restricts designers' ability to convey complex design intentions and constraints. Secondly, most existing models focus on generating labels and coordinates, while real layouts contain a range of style properties. To address these limitations, we propose a novel framework, CoLay, that integrates multiple condition types and generates complex layouts with diverse style properties. Our approach outperforms prior works in terms of generation quality and condition satisfaction while empowering users to express their design intents using a flexible combination of modalities, including natural language prompts, layout guidelines, element types, and partially completed designs.

Large language models (LLMs) have shown exceptional proficiency in natural language processing but often fall short of generating creative and original responses to open-ended questions. To enhance LLM creativity, our key insight is to emulate the human process of inducing collective creativity through engaging discussions with participants from diverse backgrounds and perspectives. To this end, we propose LLM Discussion, a three-phase discussion framework that facilitates vigorous and diverging idea exchanges and ensures convergence to creative answers. Moreover, we adopt a role-playing technique by assigning distinct roles to LLMs to combat the homogeneity of LLMs. We evaluate the efficacy of the proposed framework with the Alternative Uses Test, Similarities Test, Instances Test, and Scientific Creativity Test through both LLM evaluation and human study. Our proposed framework outperforms single-LLM approaches and existing multi-LLM frameworks across various creativity metrics.

The proliferation of LLMs in various NLP tasks has sparked debates regarding their reliability, particularly in annotation tasks where biases and hallucinations may arise. In this shared task, we address the challenge of distinguishing annotations made by LLMs from those made by human domain experts in the context of COVID-19 symptom detection from tweets in Latin American Spanish. This paper presents BrainStorm @ iREL's approach to the SMM4H 2024 Shared Task, leveraging the inherent topical information in tweets, we propose a novel approach to identify and classify annotations, aiming to enhance the trustworthiness of annotated data.

Humans read in words, but LLMs read in tokens, which are distinct units in a sentence that have consistent and significant meanings. Besides dictionary words, they also include suffixes, common expressions, names, and more. The more tokens a model encodes, the faster the model can "read" a sentence and the less computing power it consumes, thus making the response cheaper. Of the 100 results, only three of them are common enough to be used in everyday conversations; everything else consisted of words and expressions used specifically in the contexts of either gambling or pornography. The longest token, lasting 10.5 Chinese characters, literally means "_free Japanese porn video to watch."

In the summer of 2023, OpenAI created a "Superalignment" team whose goal was to steer and control future AI systems that could be so powerful they could lead to human extinction. Less than a year later, that team is dead. OpenAI told Bloomberg that the company was "integrating the group more deeply across its research efforts to help the company achieve its safety goals." But a series of tweets from Jan Leike, one of the team's leaders who recently quit revealed internal tensions between the safety team and the larger company. In a statement posted on X on Friday, Leike said that the Superalignment team had been fighting for resources to get research done.

In July last year, OpenAI announced the formation of a new research team that would prepare for the advent of supersmart artificial intelligence capable of outwitting and overpowering its creators. Ilya Sutskever, OpenAI's chief scientist and one of the company's cofounders, was named as the colead of this new team. OpenAI said the team would receive 20 percent of its computing power. Now OpenAI's "superalignment team" is no more, the company confirms. That comes after the departures of several researchers involved, Tuesday's news that Sutskever was leaving the company, and the resignation of the team's other colead.