Efficient traffic signal control is critical for reducing traffic congestion and improving overall transportation efficiency. The dynamic nature of traffic flow has prompted researchers to explore Reinforcement Learning (RL) for traffic signal control (TSC). Compared with traditional methods, RL-based solutions have shown preferable performance. However, the application of RL-based traffic signal controllers in the real world is limited by the low sample efficiency and high computational requirements of these solutions. In this work, we propose DTLight, a simple yet powerful lightweight Decision Transformer-based TSC method that can learn policy from easily accessible offline datasets. DTLight novelly leverages knowledge distillation to learn a lightweight controller from a well-trained larger teacher model to reduce implementation computation. Additionally, it integrates adapter modules to mitigate the expenses associated with fine-tuning, which makes DTLight practical for online adaptation with minimal computation and only a few fine-tuning steps during real deployment. Moreover, DTLight is further enhanced to be more applicable to real-world TSC problems. Extensive experiments on synthetic and real-world scenarios show that DTLight pre-trained purely on offline datasets can outperform state-of-the-art online RL-based methods in most scenarios. Experiment results also show that online fine-tuning further improves the performance of DTLight by up to 42.6% over the best online RL baseline methods. In this work, we also introduce Datasets specifically designed for TSC with offline RL (referred to as DTRL). Our datasets and code are publicly available.

We examine how large language models (LLMs) generalize from abstract declarative statements in their training data. As an illustration, consider an LLM that is prompted to generate weather reports for London in 2050. One possibility is that the temperatures in the reports match the mean and variance of reports from 2023 (i.e. Another possibility is that the reports predict higher temperatures, by incorporating declarative statements about climate change from scientific papers written in 2023. An example of such a declarative statement is "global temperatures will increase by 1 To test the influence of abstract declarative statements, we construct tasks in which LLMs are finetuned on both declarative and procedural information. We find that declarative statements influence model predictions, even when they conflict with procedural information. In particular, finetuning on a declarative statement S increases the model likelihood for logical consequences of S. The effect of declarative statements is consistent across three domains: aligning an AI assistant, predicting weather, and predicting demographic features. Through a series of ablations, we show that the effect of declarative statements cannot be explained by associative learning based on matching keywords. Nevertheless, the effect of declarative statements on model likelihoods is small in absolute terms and increases surprisingly little with model size (i.e. from 330 million to 175 billion parameters). We argue that these results have implications for AI risk (in relation to the "treacherous turn") and for fairness. Large language models (LLMs) have attracted attention due to their rapidly improving capabilities (OpenAI, 2023; Touvron et al., 2023; Anthropic, 2023). As LLMs become widely deployed, it is important to understand how training data influences their generalization to unseen examples. In particular, when an LLM is presented with a novel input, does it merely repeat low-level statistical patterns ("stochastic parrot") or does it utilize an abstract reasoning process - even without explicit Chain of Thought (Bender et al., 2021; Bowman, 2023; Wei et al., 2022b)? Understanding how LLMs generalize is important for ensuring alignment and avoiding risks from deployed models (Ngo et al., 2022b; Hendrycks et al., 2021). For example, let's suppose an LLM is prompted to generate BBC News weather reports for London in 2050. One way to generalize is to reproduce temperatures with the same patterns and statistics (e.g. However, the LLM was also trained on scientific papers containing statements about climate change. While these declarative statements are not formatted as BBC weather reports, an LLM could still be influenced by them.

Saturn is an innovative platform that assists Foundation Model (FM) building and its integration with IT operations (Ops). It is custom-made to meet the requirements of data scientists, enabling them to effectively create and implement FMs while enhancing collaboration within their technical domain. By offering a wide range of tools and features, Saturn streamlines and automates different stages of FM development, making it an invaluable asset for data science teams. This white paper introduces prospective applications of generative AI models derived from FMs in the financial sector.

The recent development of large language models (LLMs) with multi-billion parameters, coupled with the creation of user-friendly application programming interfaces (APIs), has paved the way for automatically generating and executing code in response to straightforward human queries. This paper explores how these emerging capabilities can be harnessed to facilitate complex scientific workflows, eliminating the need for traditional coding methods. We present initial findings from our attempt to integrate Phyloflow with OpenAI's function-calling API, and outline a strategy for developing a comprehensive workflow management system based on these concepts.

Robots are becoming increasingly popular in a wide range of environments due to their exceptional work capacity, precision, efficiency, and scalability. This development has been further encouraged by advances in Artificial Intelligence, particularly Machine Learning. By employing sophisticated neural networks, robots are given the ability to detect and interact with objects in their vicinity. However, a significant drawback arises from the underlying dependency on extensive datasets and the availability of substantial amounts of training data for these object detection models. This issue becomes particularly problematic when the specific deployment location of the robot and the surroundings, are not known in advance. The vast and ever-expanding array of objects makes it virtually impossible to comprehensively cover the entire spectrum of existing objects using preexisting datasets alone. The goal of this dissertation was to teach a robot unknown objects in the context of Human-Robot Interaction (HRI) in order to liberate it from its data dependency, unleashing it from predefined scenarios. In this context, the combination of eye tracking and Augmented Reality created a powerful synergy that empowered the human teacher to communicate with the robot and effortlessly point out objects by means of human gaze. This holistic approach led to the development of a multimodal HRI system that enabled the robot to identify and visually segment the Objects of Interest in 3D space. Through the class information provided by the human, the robot was able to learn the objects and redetect them at a later stage. Due to the knowledge gained from this HRI based teaching, the robot's object detection capabilities exhibited comparable performance to state-of-the-art object detectors trained on extensive datasets, without being restricted to predefined classes, showcasing its versatility and adaptability.

Large language models (LLMs) are increasingly used for complex tasks requiring multiple chained generation calls, advanced prompting techniques, control flow, and interaction with external environments. However, efficient systems for programming and executing these applications are lacking. To bridge this gap, we introduce SGLang, a Structured Generation Language for LLMs. SGLang is designed for the efficient programming of LLMs and incorporates primitives for common LLM programming patterns. We have implemented SGLang as a domain-specific language embedded in Python, and we developed an interpreter, a compiler, and a high-performance runtime for SGLang. These components work together to enable optimizations such as parallelism, batching, caching, sharing, and other compilation techniques. Additionally, we propose RadixAttention, a novel technique that maintains a Least Recently Used (LRU) cache of the Key-Value (KV) cache for all requests in a radix tree, enabling automatic KV cache reuse across multiple generation calls at runtime. SGLang simplifies the writing of LLM programs and boosts execution efficiency. Our experiments demonstrate that SGLang can speed up common LLM tasks by up to 5x, while reducing code complexity and enhancing control.

In the face of growing vulnerabilities found in open-source software, the need to identify {discreet} security patches has become paramount. The lack of consistency in how software providers handle maintenance often leads to the release of security patches without comprehensive advisories, leaving users vulnerable to unaddressed security risks. To address this pressing issue, we introduce a novel security patch detection system, LLMDA, which capitalizes on Large Language Models (LLMs) and code-text alignment methodologies for patch review, data enhancement, and feature combination. Within LLMDA, we initially utilize LLMs for examining patches and expanding data of PatchDB and SPI-DB, two security patch datasets from recent literature. We then use labeled instructions to direct our LLMDA, differentiating patches based on security relevance. Following this, we apply a PTFormer to merge patches with code, formulating hybrid attributes that encompass both the innate details and the interconnections between the patches and the code. This distinctive combination method allows our system to capture more insights from the combined context of patches and code, hence improving detection precision. Finally, we devise a probabilistic batch contrastive learning mechanism within batches to augment the capability of the our LLMDA in discerning security patches. The results reveal that LLMDA significantly surpasses the start of the art techniques in detecting security patches, underscoring its promise in fortifying software maintenance.

Large Language Models (LLMs) have shown remarkable performance in various basic natural language tasks. For completing the complex task, we still need a plan for the task to guide LLMs to generate the specific solutions step by step. LLMs can directly generate task plans, but these plans may still contain factual errors or are incomplete. A high-quality task plan contains correct step-by-step solutions for solving all situations and behavioral instructions for avoiding mistakes. To obtain it, we propose the Learning to Plan method, which involves two phases: (1) In the first learning task plan phase, it iteratively updates the task plan with new step-by-step solutions and behavioral instructions, which are obtained by prompting LLMs to derive from training error feedback. (2) In the subsequent test phase, the LLM uses the learned task plan to guide the inference of LLM on the test set. We demonstrate the effectiveness of our method on the five different reasoning type tasks (8 datasets). Further, our analysis experiment shows that the task plan learned by one LLM can directly guide another LLM to improve its performance, which reveals a new transfer learning paradigm. We release the code at \url{https://github.com/Eureka6174/LearnNLPlan}

Detecting protein-protein interactions (PPIs) is crucial for understanding genetic mechanisms, disease pathogenesis, and drug design. However, with the fast-paced growth of biomedical literature, there is a growing need for automated and accurate extraction of PPIs to facilitate scientific knowledge discovery. Pre-trained language models, such as generative pre-trained transformers (GPT) and bidirectional encoder representations from transformers (BERT), have shown promising results in natural language processing (NLP) tasks. We evaluated the performance of PPI identification of multiple GPT and BERT models using three manually curated gold-standard corpora: Learning Language in Logic (LLL) with 164 PPIs in 77 sentences, Human Protein Reference Database with 163 PPIs in 145 sentences, and Interaction Extraction Performance Assessment with 335 PPIs in 486 sentences. BERT-based models achieved the best overall performance, with BioBERT achieving the highest recall (91.95%) and F1-score (86.84%) and PubMedBERT achieving the highest precision (85.25%). Interestingly, despite not being explicitly trained for biomedical texts, GPT-4 achieved commendable performance, comparable to the top-performing BERT models. It achieved a precision of 88.37%, a recall of 85.14%, and an F1-score of 86.49% on the LLL dataset. These results suggest that GPT models can effectively detect PPIs from text data, offering promising avenues for application in biomedical literature mining. Further research could explore how these models might be fine-tuned for even more specialized tasks within the biomedical domain.

In a rapidly evolving digital landscape autonomous tools and robots are becoming commonplace. Recognizing the significance of this development, this paper explores the integration of Large Language Models (LLMs) like Generative pre-trained transformer (GPT) into human-robot teaming environments to facilitate variable autonomy through the means of verbal human-robot communication. In this paper, we introduce a novel framework for such a GPT-powered multi-robot testbed environment, based on a Unity Virtual Reality (VR) setting. This system allows users to interact with robot agents through natural language, each powered by individual GPT cores. By means of OpenAI's function calling, we bridge the gap between unstructured natural language input and structure robot actions. A user study with 12 participants explores the effectiveness of GPT-4 and, more importantly, user strategies when being given the opportunity to converse in natural language within a multi-robot environment. Our findings suggest that users may have preconceived expectations on how to converse with robots and seldom try to explore the actual language and cognitive capabilities of their robot collaborators. Still, those users who did explore where able to benefit from a much more natural flow of communication and human-like back-and-forth. We provide a set of lessons learned for future research and technical implementations of similar systems.