Scaling up language models to billions of parameters has opened up possibilities for in-context learning, allowing instruction tuning and few-shot learning on tasks that the model was not specifically trained for. This has achieved breakthrough performance on language tasks such as translation, summarization, and question-answering. Furthermore, in addition to these associative "System 1" tasks, recent advances in Chain-of-thought prompt learning have demonstrated strong "System 2" reasoning abilities, answering a question in the field of artificial general intelligence whether LLMs can reason. The field started with the question whether LLMs can solve grade school math word problems. This paper reviews the rapidly expanding field of prompt-based reasoning with LLMs. Our taxonomy identifies different ways to generate, evaluate, and control multi-step reasoning. We provide an in-depth coverage of core approaches and open problems, and we propose a research agenda for the near future. Finally, we highlight the relation between reasoning and prompt-based learning, and we discuss the relation between reasoning, sequential decision processes, and reinforcement learning. We find that self-improvement, self-reflection, and some metacognitive abilities of the reasoning processes are possible through the judicious use of prompts. True self-improvement and self-reasoning, to go from reasoning with LLMs to reasoning by LLMs, remains future work.

Large Language Models (LLMs) are increasingly integrated into diverse industries, posing substantial security risks due to unauthorized replication and misuse. To mitigate these concerns, robust identification mechanisms are widely acknowledged as an effective strategy. Identification systems for LLMs now rely heavily on watermarking technology to manage and protect intellectual property and ensure data security. However, previous studies have primarily concentrated on the basic principles of algorithms and lacked a comprehensive analysis of watermarking theory and practice from the perspective of intelligent identification. To bridge this gap, firstly, we explore how a robust identity recognition system can be effectively implemented and managed within LLMs by various participants using watermarking technology. Secondly, we propose a mathematical framework based on mutual information theory, which systematizes the identification process to achieve more precise and customized watermarking. Additionally, we present a comprehensive evaluation of performance metrics for LLM watermarking, reflecting participant preferences and advancing discussions on its identification applications. Lastly, we outline the existing challenges in current watermarking technologies and theoretical frameworks, and provide directional guidance to address these challenges. Our systematic classification and detailed exposition aim to enhance the comparison and evaluation of various methods, fostering further research and development toward a transparent, secure, and equitable LLM ecosystem.

Large Language Models (LLMs) are increasingly used for boosting organizational efficiency and automating tasks. While not originally designed for complex cognitive processes, recent efforts have further extended to employ LLMs in activities such as reasoning, planning, and decision-making. In business processes, such abilities could be invaluable for leveraging on the massive corpora LLMs have been trained on for gaining deep understanding of such processes. In this work, we plant the seeds for the development of a benchmark to assess the ability of LLMs to reason about causal and process perspectives of business operations. We refer to this view as Causally-augmented Business Processes (BP^C). The core of the benchmark comprises a set of BP^C related situations, a set of questions about these situations, and a set of deductive rules employed to systematically resolve the ground truth answers to these questions. Also with the power of LLMs, the seed is then instantiated into a larger-scale set of domain-specific situations and questions. Reasoning on BP^C is of crucial importance for process interventions and process improvement. Our benchmark, accessible at https://huggingface.co/datasets/ibm/BPC, can be used in one of two possible modalities: testing the performance of any target LLM and training an LLM to advance its capability to reason about BP^C.

We present Cephalo, a series of multimodal vision large language models (V-LLMs) designed for materials science applications, integrating visual and linguistic data for enhanced understanding. A key innovation of Cephalo is its advanced dataset generation method. Cephalo is trained on integrated image and text data from thousands of scientific papers and science-focused Wikipedia data demonstrates can interpret complex visual scenes, generate precise language descriptions, and answer queries about images effectively. The combination of a vision encoder with an autoregressive transformer supports multimodal natural language understanding, which can be coupled with other generative methods to create an image-to-text-to-3D pipeline. To develop more capable models from smaller ones, we report both mixture-of-expert methods and model merging. We examine the models in diverse use cases that incorporate biological materials, fracture and engineering analysis, protein biophysics, and bio-inspired design based on insect behavior. Generative applications include bio-inspired designs, including pollen-inspired architected materials, as well as the synthesis of bio-inspired material microstructures from a photograph of a solar eclipse. Additional model fine-tuning with a series of molecular dynamics results demonstrate Cephalo's enhanced capabilities to accurately predict statistical features of stress and atomic energy distributions, as well as crack dynamics and damage in materials.

Graph neural networks (GNNs) are widely applied in graph data modeling. However, existing GNNs are often trained in a task-driven manner that fails to fully capture the intrinsic nature of the graph structure, resulting in sub-optimal node and graph representations. To address this limitation, we propose a novel Graph structure Prompt Learning method (GPL) to enhance the training of GNNs, which is inspired by prompt mechanisms in natural language processing. GPL employs task-independent graph structure losses to encourage GNNs to learn intrinsic graph characteristics while simultaneously solving downstream tasks, producing higher-quality node and graph representations. In extensive experiments on eleven real-world datasets, after being trained by GPL, GNNs significantly outperform their original performance on node classification, graph classification, and edge prediction tasks (up to 10.28%, 16.5%, and 24.15%, respectively). By allowing GNNs to capture the inherent structural prompts of graphs in GPL, they can alleviate the issue of over-smooth and achieve new state-of-the-art performances, which introduces a novel and effective direction for GNN research with potential applications in various domains.

Large language models (LLMs) have recently seen widespread adoption, in both academia and industry. As these models grow, they become valuable intellectual property (IP), reflecting enormous investments by their owners. Moreover, the high cost of cloud-based deployment has driven interest towards deployment to edge devices, yet this risks exposing valuable parameters to theft and unauthorized use. Current methods to protect models' IP on the edge have limitations in terms of practicality, loss in accuracy, or suitability to requirements. In this paper, we introduce a novel hybrid inference algorithm, named SLIP, designed to protect edge-deployed models from theft. SLIP is the first hybrid protocol that is both practical for real-world applications and provably secure, while having zero accuracy degradation and minimal impact on latency. It involves partitioning the model between two computing resources, one secure but expensive, and another cost-effective but vulnerable. This is achieved through matrix decomposition, ensuring that the secure resource retains a maximally sensitive portion of the model's IP while performing a minimal amount of computations, and vice versa for the vulnerable resource. Importantly, the protocol includes security guarantees that prevent attackers from exploiting the partition to infer the secured information. Finally, we present experimental results that show the robustness and effectiveness of our method, positioning it as a compelling solution for protecting LLMs.

Automated heuristic design (AHD) has gained considerable attention for its potential to automate the development of effective heuristics. The recent advent of large language models (LLMs) has paved a new avenue for AHD, with initial efforts focusing on framing AHD as an evolutionary program search (EPS) problem. However, inconsistent benchmark settings, inadequate baselines, and a lack of detailed component analysis have left the necessity of integrating LLMs with search strategies and the true progress achieved by existing LLM-based EPS methods to be inadequately justified. This work seeks to fulfill these research queries by conducting a large-scale benchmark comprising four LLM-based EPS methods and four AHD problems across nine LLMs and five independent runs. Our extensive experiments yield meaningful insights, providing empirical grounding for the importance of evolutionary search in LLM-based AHD approaches, while also contributing to the advancement of future EPS algorithmic development. To foster accessibility and reproducibility, we have fully open-sourced our benchmark and corresponding results.

The increasing consumption of news online in the 21st century coincided with increased publication of disinformation, biased reporting, hate speech and other unwanted Web content. We describe BiasScanner, an application that aims to strengthen democracy by supporting news consumers with scrutinizing news articles they are reading online. BiasScanner contains a server-side pre-trained large language model to identify biased sentences of news articles and a front-end Web browser plug-in. At the time of writing, BiasScanner can identify and classify more than two dozen types of media bias at the sentence level, making it the most fine-grained model and only deployed application (automatic system in use) of its kind. It was implemented in a light-weight and privacy-respecting manner, and in addition to highlighting likely biased sentence it also provides explanations for each classification decision as well as a summary analysis for each news article. While prior research has addressed news bias detection, we are not aware of any work that resulted in a deployed browser plug-in (c.f. also biasscanner.org for a Web demo).

In academic research, systematic literature reviews are foundational and highly relevant, yet tedious to create due to the high volume of publications and labor-intensive processes involved. Systematic selection of relevant papers through conventional means like keyword-based filtering techniques can sometimes be inadequate, plagued by semantic ambiguities and inconsistent terminology, which can lead to sub-optimal outcomes. To mitigate the required extensive manual filtering, we explore and evaluate the potential of using Large Language Models (LLMs) to enhance the efficiency, speed, and precision of literature review filtering, reducing the amount of manual screening required. By using models as classification agents acting on a structured database only, we prevent common problems inherent in LLMs, such as hallucinations. We evaluate the real-world performance of such a setup during the construction of a recent literature survey paper with initially more than 8.3k potentially relevant articles under consideration and compare this with human performance on the same dataset. Our findings indicate that employing advanced LLMs like GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Flash, or Llama3 with simple prompting can significantly reduce the time required for literature filtering - from usually weeks of manual research to only a few minutes. Simultaneously, we crucially show that false negatives can indeed be controlled through a consensus scheme, achieving recalls >98.8% at or even beyond the typical human error threshold, thereby also providing for more accurate and relevant articles selected. Our research not only demonstrates a substantial improvement in the methodology of literature reviews but also sets the stage for further integration and extensive future applications of responsible AI in academic research practices.

A language is made up of an infinite/finite number of sentences, which in turn is composed of a number of words. The Electrocardiogram (ECG) is the most popular noninvasive medical tool for studying heart function and diagnosing various irregular cardiac rhythms. Intuitive inspection of the ECG reveals a marked similarity between ECG signals and the spoken language. As a result, the ECG signal may be thought of as a series of heartbeats (similar to sentences in a spoken language), with each heartbeat consisting of a collection of waves (similar to words in a sentence) with varying morphologies. Just as natural language processing (NLP) is used to help computers comprehend and interpret human natural language, it is conceivable to create NLP-inspired algorithms to help computers comprehend the electrocardiogram data more efficiently. In this study, we propose a novel ECG analysis technique, based on embedding and self attention, to capture the spatial as well as the temporal dependencies of the ECG data. To generate the embedding, an encoder-decoder network was proposed to capture the temporal dependencies of the ECG signal and perform data compression. The compressed and encoded data was fed to the embedding layer as its weights. Finally, the proposed CNN-LSTM-Self Attention classifier works on the embedding layer and classifies the signal as normal or anomalous. The approach was tested using the PTB-xl dataset, which is severely imbalanced. Our emphasis was to appropriately recognise the disease classes present in minority numbers, in order to limit the detection of False Negative cases. An accuracy of 91% was achieved with a good F1-score for all the disease classes. Additionally, the the size of the model was reduced by 34% due to compression, making it suitable for deployment in real time applications