Ranking models play a crucial role in enhancing overall accuracy of text retrieval systems. These multi-stage systems typically utilize either dense embedding models or sparse lexical indices to retrieve relevant passages based on a given query, followed by ranking models that refine the ordering of the candidate passages by its relevance to the query. This paper benchmarks various publicly available ranking models and examines their impact on ranking accuracy. We focus on text retrieval for question-answering tasks, a common use case for Retrieval-Augmented Generation systems. Our evaluation benchmarks include models some of which are commercially viable for industrial applications. We introduce a state-of-the-art ranking model, NV-RerankQA-Mistral-4B-v3, which achieves a significant accuracy increase of ~14% compared to pipelines with other rerankers. We also provide an ablation study comparing the fine-tuning of ranking models with different sizes, losses and self-attention mechanisms. Finally, we discuss challenges of text retrieval pipelines with ranking models in real-world industry applications, in particular the trade-offs among model size, ranking accuracy and system requirements like indexing and serving latency / throughput.

Federated instruction tuning enables multiple clients to collaboratively fine-tune a shared large language model (LLM) that can follow humans' instructions without directly sharing raw data. However, existing literature impractically requires that all the clients readily hold instruction-tuning data (i.e., structured instruction-response pairs), which necessitates massive human annotations since clients' data is usually unstructured text instead. Addressing this, we propose a novel and flexible framework FedIT-U2S, which can automatically transform unstructured corpus into structured data for federated instruction tuning. FedIT-U2S consists two key steps: (1) few-shot instruction-tuning data generation, where each unstructured data piece together with several examples is combined to prompt an LLM in generating an instruction-response pair. To further enhance the flexibility, a retrieval-based example selection technique is proposed, where the examples are automatically selected based on the relatedness between the client's data piece and example pool, bypassing the need of determining examples in advance. (2) A typical federated instruction tuning process based on the generated data. Overall, FedIT-U2S can be applied to diverse scenarios as long as the client holds valuable text corpus, broadening the application scope of federated instruction tuning. We conduct a series of experiments on three domains (medicine, knowledge, and math), showing that our proposed FedIT-U2S can consistently and significantly brings improvement over the base LLM.

Speech language models have recently demonstrated great potential as universal speech processing systems. Such models have the ability to model the rich acoustic information existing in audio signals, beyond spoken content, such as emotion, background noise, etc. Despite this, evaluation benchmarks which evaluate awareness to a wide range of acoustic aspects, are lacking. To help bridge this gap, we introduce SALMon, a novel evaluation suite encompassing background noise, emotion, speaker identity and room impulse response. The proposed benchmarks both evaluate the consistency of the inspected element and how much it matches the spoken text. We follow a modelling based approach, measuring whether a model gives correct samples higher scores than incorrect ones. This approach makes the benchmark fast to compute even for large models. We evaluated several speech language models on SALMon, thus highlighting the strengths and weaknesses of each evaluated method. Code and data are publicly available at https://pages.cs.huji.ac.il/adiyoss-lab/salmon/ .

Multimodal Large Language Models (MLLMs) are distinguished by their multimodal comprehensive ability and widely used in many real-world applications including GPT-4o, autonomous driving and robotics. Despite their impressive performance, the multimodal inputs always incur long context. The inference under long context requires caching massive Key and Value states (KV cache) of previous tokens, which introduces high latency and excessive memory consumption. Due to this reason, it is challenging to deploy streaming inference of MLLMs on edge devices, which largely constrains the power and usage of MLLMs in real-world applications. In this paper, we introduce Inf-MLLM, an efficient inference framework for MLLMs, which enable streaming inference of MLLM on a single GPU with infinite context. Inf-MLLM is based on our key observation of the attention pattern in both LLMs and MLLMs called "attention saddles". Thanks to the newly discovered attention pattern, Inf-MLLM maintains a size-constrained KV cache by dynamically caching recent tokens and relevant tokens. Furthermore, Inf-MLLM proposes attention bias, a novel approach to enable MLLMs to capture long-term dependency. We show that Inf-MLLM enables multiple LLMs and MLLMs to achieve stable performance over 4M-token long texts and multi-round conversations with 1-hour-long videos on a single GPU. In addition, Inf-MLLM exhibits superior streaming reasoning quality than existing methods such as StreamingLLM and 2x speedup than H2O.

There have been growing concerns around high-stake applications For decades, medicine has been marred by implicit and explicit that rely on models trained with biased data, which consequently biases that continue to negatively impact patient outcomes by perpetuating produce biased predictions, often harming the most vulnerable. In stereotypes and contributing to health disparities among particular, biased medical data could cause health-related applications social groups that face systemic oppression [8, 9]. Despite efforts to and recommender systems to create outputs that jeopardize remediate and address these biases from their source, many medical patient care and widen disparities in health outcomes. A recent schools still incorporate biased medical teachings during the framework titled Fairness via AI posits that, instead of attempting preclinical years [12, 28]. Many educators continue to misuse race to correct model biases, researchers must focus on their root causes as a substitute for genetics or ancestry, or they use gender and sex by using AI to debias data. Inspired by this framework, we tackle terms incorrectly reinforcing the notion that sex and gender are bias detection in medical curricula using NLP models, including binary or fixed rather than fluid, which can potentially alienate LLMs, and evaluate them on a gold standard dataset containing gender-nonconforming students and patients [1, 14, 15]. The current 4,105 excerpts annotated by medical experts for bias from a large focus in AI research is primarily on identifying and exposing corpus. We build on previous work by coauthors which augments bias within AI systems, often without addressing the root causes the set of negative samples with non-annotated text containing social of bias inherent in the data these systems are built upon.

We introduce SimulBench, a benchmark designed to evaluate large language models (LLMs) across a diverse collection of creative simulation scenarios, such as acting as a Linux terminal or playing text games with users. While these simulation tasks serve as effective measures of an LLM's general intelligence, they are seldom incorporated into existing benchmarks. A major challenge is to develop an evaluation framework for testing different LLMs fairly while preserving the multi-round interactive nature of simulation tasks between users and AI. To tackle this issue, we suggest using a fixed LLM as a user agent to engage with an LLM to collect dialogues first under different tasks. Then, challenging dialogue scripts are extracted for evaluating different target LLMs. To facilitate automatic assessment on \DataName{}, GPT-4 is employed as the evaluator, tasked with reviewing the quality of the final response generated by the target LLMs given multi-turn dialogue scripts. Our comprehensive experiments indicate that these simulation tasks continue to pose a significant challenge with their unique natures and show the gap between proprietary models and the most advanced open LLMs. For example, GPT-4-turbo outperforms LLaMA-3-70b-Chat on 18.55\% more cases.

Interactive feedback, where feedback flows in both directions between teacher and student, is more effective than traditional one-way feedback. However, it is often too time-consuming for widespread use in educational practice. While Large Language Models (LLMs) have potential for automating feedback, they struggle with reasoning and interaction in an interactive setting. This paper introduces CAELF, a Contestable AI Empowered LLM Framework for automating interactive feedback. CAELF allows students to query, challenge, and clarify their feedback by integrating a multi-agent system with computational argumentation. Essays are first assessed by multiple Teaching-Assistant Agents (TA Agents), and then a Teacher Agent aggregates the evaluations through formal reasoning to generate feedback and grades. Students can further engage with the feedback to refine their understanding. A case study on 500 critical thinking essays with user studies demonstrates that CAELF significantly improves interactive feedback, enhancing the reasoning and interaction capabilities of LLMs. This approach offers a promising solution to overcoming the time and resource barriers that have limited the adoption of interactive feedback in educational settings.

Multimodal Large Language Models (MLLMs) have demonstrated great zero-shot performance on visual question answering (VQA). However, when it comes to knowledge-based VQA (KB-VQA), MLLMs may lack human commonsense or specialized domain knowledge to answer such questions and require obtaining necessary information from external knowledge sources. Previous works like Retrival-Augmented VQA-v2 (RAVQA-v2) focus on utilizing as much input information, such as image-based textual descriptions and retrieved knowledge, as possible to improve performance, but they all overlook the issue that with the number of input tokens increasing, inference efficiency significantly decreases, which contradicts the demands of practical applications. To address this issue, we propose Retrieval-Augmented MLLM with Compressed Contexts (RACC). RACC learns to compress and aggregate retrieved contexts, from which it generates a compact modulation in the form of Key-Value (KV) cache. This modulation is then used to adapt the downstream frozen MLLM, thereby achieving effective and efficient inference. RACC achieves a state-of-the-art (SOTA) performance of 62.9% on OK-VQA. Moreover, it significantly reduces inference latency by 22.0%-59.7% compared to the prominent RAVQA-v2. Abundant experiments show RACC's broad applicability. It is compatible with various off-the-shelf MLLMs and can also handle different knowledge sources including textual and multimodal documents.

The vast pre-existing slides serve as rich and important materials to carry lecture knowledge. However, effectively leveraging lecture slides to serve students is difficult due to the multi-modal nature of slide content and the heterogeneous teaching actions. We study the problem of discovering effective designs that convert a slide into an interactive lecture. We develop Slide2Lecture, a tuning-free and knowledge-regulated intelligent tutoring system that can (1) effectively convert an input lecture slide into a structured teaching agenda consisting of a set of heterogeneous teaching actions; (2) create and manage an interactive lecture that generates responsive interactions catering to student learning demands while regulating the interactions to follow teaching actions. Slide2Lecture contains a complete pipeline for learners to obtain an interactive classroom experience to learn the slide. For teachers and developers, Slide2Lecture enables customization to cater to personalized demands. The evaluation rated by annotators and students shows that Slide2Lecture is effective in outperforming the remaining implementation. Slide2Lecture's online deployment has made more than 200K interaction with students in the 3K lecture sessions. We open source Slide2Lecture's implementation in https://anonymous.4open.science/r/slide2lecture-4210/.

Conversational AI (CAI) systems that use Large Language Models (LLMs) such as GPT (Generative Pre-trained Transformer) have been shown to be fruitful in augmenting human creativity, providing numerous novel and diverse ideas. Despite the success in ideation quantity, the qualitative assessment of these ideas remains challenging and traditionally reliant on expert human evaluation. This method suffers from limitations such as human judgment errors, bias, and oversight. Addressing this gap, our study introduces a comprehensive mathematical framework for automated analysis to objectively evaluate the plethora of ideas generated by CAI systems and/or humans. This framework is particularly advantageous for novice designers who lack experience in selecting promising ideas. By converting the ideas into higher dimensional vectors and quantitatively measuring the diversity between them using tools such as UMAP, DBSCAN and PCA, the proposed method provides a reliable and objective way of selecting the most promising ideas, thereby enhancing the efficiency of the ideation phase.