We propose a modular framework that leverages the expertise of different foundation models over different modalities and domains in order to perform a single, complex, multi-modal task, without relying on prompt engineering or otherwise tailor-made multi-modal training. Our approach enables decentralized command execution and allows each model to both contribute and benefit from the expertise of the other models. Our method can be extended to a variety of foundation models (including audio and vision), above and beyond only language models, as it does not depend on prompts. We demonstrate our approach on two tasks. On the well-known task of stylized image captioning, our experiments show that our approach outperforms semi-supervised state-of-the-art models, while being zero-shot and avoiding costly training, data collection, and prompt engineering. We further demonstrate this method on a novel task, audio-aware image captioning, in which an image and audio are given and the task is to generate text that describes the image within the context of the provided audio. Our code is available on GitHub.

Detecting factual errors in textual information, whether generated by large language models (LLM) or curated by humans, is crucial for making informed decisions. LLMs' inability to attribute their claims to external knowledge and their tendency to hallucinate makes it difficult to rely on their responses. Humans, too, are prone to factual errors in their writing. Since manual detection and correction of factual errors is labor-intensive, developing an automatic approach can greatly reduce human effort. We present FLEEK, a prototype tool that automatically extracts factual claims from text, gathers evidence from external knowledge sources, evaluates the factuality of each claim, and suggests revisions for identified errors using the collected evidence. Initial empirical evaluation on fact error detection (77-85\% F1) shows the potential of FLEEK. A video demo of FLEEK can be found at https://youtu.be/NapJFUlkPdQ.

In an era marked by the increasing adoption of Large Language Models (LLMs) for various tasks, there is a growing focus on exploring LLMs' capabilities in handling web data, particularly graph data. Dynamic graphs, which capture temporal network evolution patterns, are ubiquitous in real-world web data. Evaluating LLMs' competence in understanding spatial-temporal information on dynamic graphs is essential for their adoption in web applications, which remains unexplored in the literature. In this paper, we bridge the gap via proposing to evaluate LLMs' spatial-temporal understanding abilities on dynamic graphs, to the best of our knowledge, for the first time. Specifically, we propose the LLM4DyG benchmark, which includes nine specially designed tasks considering the capability evaluation of LLMs from both temporal and spatial dimensions. Then, we conduct extensive experiments to analyze the impacts of different data generators, data statistics, prompting techniques, and LLMs on the model performance. Finally, we propose Disentangled Spatial-Temporal Thoughts (DST2) for LLMs on dynamic graphs to enhance LLMs' spatial-temporal understanding abilities. Our main observations are: 1) LLMs have preliminary spatial-temporal understanding abilities on dynamic graphs, 2) Dynamic graph tasks show increasing difficulties for LLMs as the graph size and density increase, while not sensitive to the time span and data generation mechanism, 3) the proposed DST2 prompting method can help to improve LLMs' spatial-temporal understanding abilities on dynamic graphs for most tasks. The data and codes will be open-sourced at publication time.

As artificial intelligence (AI) gains greater adoption in a wide variety of applications, it has immense potential to contribute to mathematical discovery, by guiding conjecture generation, constructing counterexamples, assisting in formalizing mathematics, and discovering connections between different mathematical areas, to name a few. While prior work has leveraged computers for exhaustive mathematical proof search, recent efforts based on large language models (LLMs) aspire to position computing platforms as co-contributors in the mathematical research process. Despite their current limitations in logic and mathematical tasks, there is growing interest in melding theorem proving systems with foundation models. This work investigates the applicability of LLMs in formalizing advanced mathematical concepts and proposes a framework that can critically review and check mathematical reasoning in research papers. Given the noted reasoning shortcomings of LLMs, our approach synergizes the capabilities of proof assistants, specifically PVS, with LLMs, enabling a bridge between textual descriptions in academic papers and formal specifications in PVS. By harnessing the PVS environment, coupled with data ingestion and conversion mechanisms, we envision an automated process, called \emph{math-PVS}, to extract and formalize mathematical theorems from research papers, offering an innovative tool for academic review and discovery.

Large language models (LLMs) such as GPT-3 have demonstrated a strong capability to generate coherent and contextually relevant text. However, amidst their successes, a crucial issue persists: their generated outputs still lack commonsense at times. Moreover, fine-tuning the entire LLM towards more commonsensical outputs is computationally expensive if not infeasible. In this paper, we present a computation-efficient framework that steers a frozen Pre-Trained Language Model (PTLM) towards more commonsensical generation (i.e., producing a plausible output that incorporates a list of concepts in a meaningful way). Specifically, we first construct a reference-free evaluator that assigns a sentence with a commonsensical score by grounding the sentence to a dynamic commonsense knowledge base from four different relational aspects. We then use the scorer as the oracle for commonsense knowledge, and extend the controllable generation method called NADO to train an auxiliary head that guides a fixed PTLM to better satisfy the oracle. We test our framework on a series of GPT-2-, Flan-T5-, and Alpaca-based language models (LMs) on two constrained concept-to-sentence benchmarks. Human evaluation results demonstrate that our method consistently leads to the most commonsensical outputs.

This paper combines two contributions. First, we introduce an extension of the Meta-World benchmark, which we call "Language-World," which allows a large language model to operate in a simulated robotic environment using semi-structured natural language queries and scripted skills described using natural language. By using the same set of tasks as Meta-World, Language-World results can be easily compared to Meta-World results, allowing for a point of comparison between recent methods using Large Language Models (LLMs) and those using Deep Reinforcement Learning. Second, we introduce a method we call Plan Conditioned Behavioral Cloning (PCBC), that allows finetuning the behavior of high-level plans using end-to-end demonstrations. Using Language-World, we show that PCBC is able to achieve strong performance in a variety of few-shot regimes, often achieving task generalization with as little as a single demonstration. We have made Language-World available as open-source software at https://github.com/krzentner/language-world/.

With the rise of generative pre-trained transformer models such as GPT-3, GPT-NeoX, or OPT, distinguishing human-generated texts from machine-generated ones has become important. We refined five separate language models to generate synthetic tweets, uncovering that shallow learning classification algorithms, like Naive Bayes, achieve detection accuracy between 0.6 and 0.8. Shallow learning classifiers differ from human-based detection, especially when using higher temperature values during text generation, resulting in a lower detection rate. Humans prioritize linguistic acceptability, which tends to be higher at lower temperature values. In contrast, transformer-based classifiers have an accuracy of 0.9 and above. We found that using a reinforcement learning approach to refine our generative models can successfully evade BERT-based classifiers with a detection accuracy of 0.15 or less.

Positioned between pre-training and user deployment, aligning large language models (LLMs) through reinforcement learning (RL) has emerged as a prevailing strategy for training instruction following-models such as ChatGPT. In this work, we initiate the study of privacy-preserving alignment of LLMs through Differential Privacy (DP) in conjunction with RL. Following the influential work of Ziegler et al. (2020), we study two dominant paradigms: (i) alignment via RL without human in the loop (e.g., positive review generation) and (ii) alignment via RL from human feedback (RLHF) (e.g., summarization in a human-preferred way). We give a new DP framework to achieve alignment via RL, and prove its correctness. Our experimental results validate the effectiveness of our approach, offering competitive utility while ensuring strong privacy protections.

As large language models (LLMs) are widely adopted, new safety issues and policies emerge, to which existing safety classifiers do not generalize well. If we have only observed a few examples of violations of a new safety rule, how can we build a classifier to detect violations? In this paper, we study the novel setting of domain-generalized few-shot learning for LLM-based text safety classifiers. Unlike prior few-shot work, these new safety issues can be hard to uncover and we do not get to choose the few examples. We demonstrate that existing few-shot techniques do not perform well in this setting, and rather we propose to do parameter-efficient fine-tuning (PEFT) combined with augmenting training data based on similar examples in prior existing rules. We empirically show that our approach of similarity-based data-augmentation + prompt-tuning (DAPT) consistently outperforms baselines that either do not rely on data augmentation or on PEFT by 7-17% F1 score in the Social Chemistry moral judgement and 9-13% AUC in the Toxicity detection tasks, even when the new rule is loosely correlated with existing ones.

Transformer-based large language models have remarkable potential to accelerate design optimization for applications such as drug development and materials discovery. Self-supervised pretraining of transformer models requires large-scale datasets, which are often sparsely populated in topical areas such as polymer science. State-of-the-art approaches for polymers conduct data augmentation to generate additional samples but unavoidably incurs extra computational costs. In contrast, large-scale open-source datasets are available for small molecules and provide a potential solution to data scarcity through transfer learning. In this work, we show that using transformers pretrained on small molecules and fine-tuned on polymer properties achieve comparable accuracy to those trained on augmented polymer datasets for a series of benchmark prediction tasks.