The non-profit settlement sector in Canada supports newcomers in achieving successful integration. This sector faces increasing operational pressures amidst rising immigration targets, which highlights a need for enhanced efficiency and innovation, potentially through reliable AI solutions. The ad-hoc use of general-purpose generative AI, such as ChatGPT, might become a common practice among newcomers and service providers to address this need. However, these tools are not tailored for the settlement domain and can have detrimental implications for immigrants and refugees. We explore the risks that these tools might pose on newcomers to first, warn against the unguarded use of generative AI, and second, to incentivize further research and development in creating AI literacy programs as well as customized LLMs that are aligned with the preferences of the impacted communities. Crucially, such technologies should be designed to integrate seamlessly into the existing workflow of the settlement sector, ensuring human oversight, trustworthiness, and accountability.

These studies have demonstrated how model performance is influenced by both the size of the model and the amount of training computation. However, there is limited knowledge on how varying the compute during inference affects model performance after the model has been trained. To improve the task performance of large language models (LLMs), inference techniques typically involve additional computation as a performance maximization step at inference time [Nye et al., 2021, Wei et al., 2022, Wang et al., 2022b, Yao et al., 2023, Chen et al., 2024b]. This cost must be taken into account for compute-optimal inference. For example, a Monte Carlo Tree Search (MCTS) method [Jones, 2021] may improve task performance, but potentially require much more compute than simply sampling solutions multiple times. Generally speaking, we need a comprehensive understanding of how various inference-time methods (e.g., Best-of-N, Majority Voting) trade off between performance and cost. To improve our understanding, this paper presents a thorough empirical evaluation with careful analysis over various configurations of representative LLMs and inference algorithms. Specifically, we explore how to select an optimal size for the language model and an effective inference strategy (e.g., Greedy Search, Majority Voting, Best-of-N, Weighted Voting, and their Tree Search variants) to maximize performance (i.e., accuracy) with a given compute budget.

Also, coherence is essential for data (Koshorek et al., 2018; Arnold et al., 2019), understanding logical structures and semantics. Enhancing contemporary supervised models (Lukasik et al., coherence modeling has achieved significant 2020; Somasundaran et al., 2020; Zhang et al., improvements in long document topic segmentation 2021; Yu et al., 2023) have demonstrated superior (Yu et al., 2023). Therefore, we improve performance compared to unsupervised approaches supervised VTS methods by thoroughly exploring (Riedl and Biemann, 2012; Solbiati et al., multimodal fusion and multimodal coherence 2021). Notably, supervised models that excel at modeling. We enhance multimodal fusion modeling long sequences (Zhang et al., 2021; Yu from the perspectives of model architecture and et al., 2023) are capable of capturing longer contextual pre-training and fine-tuning tasks. Specifically, we nuances and thereby achieve better topic segmentation compare various multimodal fusion architectures performance, compared to models that built upon Cross-Attention and Mixture-of-Experts model local sentence pairs or block pairs (Wang (MoE). We investigate the effect of multi-modal et al., 2017; Lukasik et al., 2020). In addition, contrastive learning for general pre-training and recent works (Somasundaran et al., 2020; Xing fine-tuning for strengthening cross-modal alignment.

The increase in open-source availability of Large Language Models (LLMs) has enabled users to deploy them on more and more resource-constrained edge devices to reduce reliance on network connections and provide more privacy. However, the high computation and memory demands of LLMs make their execution on resource-constrained edge devices challenging and inefficient. To address this issue, designing new and efficient edge accelerators for LLM inference is crucial. FPGA-based accelerators are ideal for LLM acceleration due to their reconfigurability, as they enable model-specific optimizations and higher performance per watt. However, creating and integrating FPGA-based accelerators for LLMs (particularly on edge devices) has proven challenging, mainly due to the limited hardware design flows for LLMs in existing FPGA platforms. To tackle this issue, in this paper we first propose a new design platform, named SECDA-LLM, that utilizes the SECDA methodology to streamline the process of designing, integrating, and deploying efficient FPGA-based LLM accelerators for the llama.cpp inference framework. We then demonstrate, through a case study, the potential benefits of SECDA-LLM by creating a new MatMul accelerator that supports block floating point quantized operations for LLMs. Our initial accelerator design, deployed on the PYNQ-Z1 board, reduces latency 1.7 seconds per token or ~2 seconds per word) by 11x over the dual-core Arm NEON-based CPU execution for the TinyLlama model.

Recently, large language models (LLMs) have been gaining a lot of interest due to their adaptability and extensibility in emerging applications, including communication networks. It is anticipated that 6G mobile edge computing networks will be able to support LLMs as a service, as they provide ultra reliable low-latency communications and closed loop massive connectivity. However, LLMs are vulnerable to data and model privacy issues that affect the trustworthiness of LLMs to be deployed for user-based services. In this paper, we explore the security vulnerabilities associated with fine-tuning LLMs in 6G networks, in particular the membership inference attack. We define the characteristics of an attack network that can perform a membership inference attack if the attacker has access to the fine-tuned model for the downstream task. We show that the membership inference attacks are effective for any downstream task, which can lead to a personal data breach when using LLM as a service. The experimental results show that the attack success rate of maximum 92% can be achieved on named entity recognition task. Based on the experimental analysis, we discuss possible defense mechanisms and present possible research directions to make the LLMs more trustworthy in the context of 6G networks.

Traditional legal retrieval systems designed to retrieve legal documents, statutes, precedents, and other legal information are unable to give satisfactory answers due to lack of semantic understanding of specific questions. Large Language Models (LLMs) have achieved excellent results in a variety of natural language processing tasks, which inspired us that we train a LLM in the legal domain to help legal retrieval. However, in the Chinese legal domain, due to the complexity of legal questions and the rigour of legal articles, there is no legal large model with satisfactory practical application yet. In this paper, we present DeliLaw, a Chinese legal counselling system based on a large language model. DeliLaw integrates a legal retrieval module and a case retrieval module to overcome the model hallucination. Users can consult professional legal questions, search for legal articles and relevant judgement cases, etc. on the DeliLaw system in a dialogue mode. In addition, DeliLaw supports the use of English for counseling. we provide the address of the system: https://data.delilegal.com/lawQuestion.

Multimodal language models (MLLMs) are increasingly being implemented in real-world environments, necessitating their ability to interpret 3D spaces and comprehend temporal dynamics. Despite their potential, current top models within our community still fall short in adequately understanding spatial and temporal dimensions. We introduce Coarse Correspondence, a simple, training-free, effective, and general-purpose visual prompting method to elicit 3D and temporal understanding in multimodal LLMs. Our method uses a lightweight tracking model to find object correspondences between frames in a video or between sets of image viewpoints. It selects the most frequent object instances and visualizes them with markers with unique IDs in the image. With this simple approach, we achieve state-of-the-art results on 3D understanding benchmarks including ScanQA (+20.5\%) and a subset of OpenEQA (+9.7\%), and on long-form video benchmarks such as EgoSchema (+6.0\%). We also curate a small diagnostic dataset to evaluate whether MLLMs can reason about space from a described viewpoint other than the camera viewpoint. Again, Coarse Correspondence improves spatial perspective-taking abilities but we highlight that MLLMs struggle with this task. Together, we demonstrate that our simple prompting method can significantly aid downstream tasks that require 3D or temporal reasoning.

Synthetically created Cross-Lingual Summarisation (CLS) datasets are prone to include document-summary pairs where the reference summary is unfaithful to the corresponding document as it contains content not supported by the document (i.e., hallucinated content). This low data quality misleads model learning and obscures evaluation results. Automatic ways to assess hallucinations and improve training have been proposed for monolingual summarisation, predominantly in English. For CLS, we propose to use off-the-shelf cross-lingual Natural Language Inference (X-NLI) to evaluate faithfulness of reference and model generated summaries. Then, we study training approaches that are aware of faithfulness issues in the training data and propose an approach that uses unlikelihood loss to teach a model about unfaithful summary sequences. Our results show that it is possible to train CLS models that yield more faithful summaries while maintaining comparable or better informativess.

Rebuses are puzzles requiring constrained multi-step reasoning to identify a hidden phrase from a set of images and letters. In this work, we introduce a large collection of verbalized rebuses for the Italian language and use it to assess the rebus-solving capabilities of state-of-the-art large language models. While general-purpose systems such as LLaMA-3 and GPT-4o perform poorly on this task, ad-hoc fine-tuning seems to improve models' performance. However, we find that performance gains from training are largely motivated by memorization. Our results suggest that rebus solving remains a challenging test bed to evaluate large language models' linguistic proficiency and sequential instruction-following skills.

MM-Vet, with open-ended vision-language questions targeting at evaluating integrated capabilities, has become one of the most popular benchmarks for large multimodal model evaluation. MM-Vet assesses six core vision-language (VL) capabilities: recognition, knowledge, spatial awareness, language generation, OCR, and math. However, its question format is restricted to single image-text pairs, lacking the interleaved image and text sequences prevalent in real-world scenarios. To address this limitation, we introduce MM-Vet v2, which includes a new VL capability called "image-text sequence understanding", evaluating models' ability to process VL sequences. Furthermore, we maintain the high quality of evaluation samples while further expanding the evaluation set size. Using MM-Vet v2 to benchmark large multimodal models, we found that Claude 3.5 Sonnet is the best model with a score of 71.8, slightly outperforming GPT-4o which scored 71.0. Among open-weight models, InternVL2-Llama3-76B leads with a score of 68.4.