The evaluation of the problem-solving capability under incomplete information scenarios of Large Language Models (LLMs) is increasingly important, encompassing capabilities such as questioning, knowledge search, error detection, and path planning. Current research mainly focus on LLMs' problem-solving capability such as ``Twenty Questions''. However, these kinds of games do not require recognizing misleading cues which are necessary in the incomplete information scenario. Moreover, the existing game such as ``Who is undercover'' are highly subjective, making it challenging for evaluation. Therefore, in this paper, we introduce a novel game named BrainKing based on the ``Who is undercover'' and ``Twenty Questions'' for evaluating LLM capabilities under incomplete information scenarios. It requires LLMs to identify target entities with limited yes-or-no questions and potential misleading answers. By setting up easy, medium, and hard difficulty modes, we comprehensively assess the performance of LLMs across various aspects. Our results reveal the capabilities and limitations of LLMs in BrainKing, providing significant insights of LLM problem-solving levels.

Contextual bandits have become an increasingly popular solution for personalized recommender systems. Despite their growing use, the interpretability of these systems remains a significant challenge, particularly for the often non-expert operators tasked with ensuring their optimal performance. In this paper, we address this challenge by designing a new interface to explain to domain experts the underlying behaviour of a bandit. Central is a metric we term "value gain", a measure derived from off-policy evaluation to quantify the real-world impact of sub-components within a bandit. We conduct a qualitative user study to evaluate the effectiveness of our interface. Our findings suggest that by carefully balancing technical rigour with accessible presentation, it is possible to empower non-experts to manage complex machine learning systems. We conclude by outlining guiding principles that other researchers should consider when building similar such interfaces in future.

Taiwan Semiconductor Manufacturing Co. and Samsung Electronics have discussed building major new factories in the United Arab Emirates in coming years to help satisfy soaring demand for artificial intelligence computing, the Wall Street Journal reported. Executives from TSMC, the world's largest chipmaker, have visited the UAE recently to discuss building a plant complex that could rival the company's advanced facilities in Taiwan, the newspaper said Sunday, citing people familiar with the interactions. South Korea's Samsung has also sent emissaries to the Middle Eastern country recently to talk about major new operations there, the Journal said, citing separate people with knowledge of the company's strategy.

Most recommender systems research focuses on binary historical user-item interaction encodings to predict future interactions. User features, item features, and interaction strengths remain largely under-utilized in this space or only indirectly utilized, despite proving largely effective in large-scale production recommendation systems. We propose a new attention mechanism, loosely based on the principles of collaborative filtering, called Row-Column Separable Attention RCSA to take advantage of real-valued interaction weights as well as user and item features directly. Building on this mechanism, we additionally propose a novel Graph Diffusion Transformer GDiT architecture which is trained to iteratively denoise the weighted interaction matrix of the user-item interaction graph directly. The weighted interaction matrix is built from the bipartite structure of the user-item interaction graph and corresponding edge weights derived from user-item rating interactions. Inspired by the recent progress in text-conditioned image generation, our method directly produces user-item rating predictions on the same scale as the original ratings by conditioning the denoising process on user and item features with a principled approach.

Current instruction-tuned language models are exclusively trained with textual preference data and thus are often not aligned with the unique requirements of other modalities, such as speech. To better align language models with the speech domain, we explore (i) prompting strategies grounded in radio-industry best practices and (ii) preference learning using a novel speech-based preference data of 20K samples, generated with a wide spectrum of prompts that induce varying dimensions of speech-suitability and labeled by annotators who listen to response pairs. Both human and automatic evaluation show that both prompting and preference learning increase the speech-suitability of popular instruction-tuned LLMs. Interestingly, we find that prompting and preference learning can be additive; combining them achieves the best win rates in head-to-head comparison, resulting in responses that are preferred or tied to the base model in 76.2% of comparisons on average. Lastly, we share lexical, syntactical, and qualitative analyses to showcase how each method contributes to improving the speech-suitability of generated responses.

Social networks form a valuable source of world knowledge, where influential entities correspond to popular accounts. Unlike factual knowledge bases (KBs), which maintain a semantic ontology, structured semantic information is not available on social media. In this work, we consider a social KB of roughly 200K popular Twitter accounts, which denotes entities of interest. We elicit semantic information about those entities. In particular, we associate them with a fine-grained set of 136 semantic types, e.g., determine whether a given entity account belongs to a politician, or a musical artist. In the lack of explicit type information in Twitter, we obtain semantic labels for a subset of the accounts via alignment with the KBs of DBpedia and Wikidata. Given the labeled dataset, we finetune a transformer-based text encoder to generate semantic embeddings of the entities based on the contents of their accounts. We then exploit this evidence alongside network-based embeddings to predict the entities semantic types. In our experiments, we show high type prediction performance on the labeled dataset. Consequently, we apply our type classification model to all of the entity accounts in the social KB. Our analysis of the results offers insights about the global semantics of the Twitter sphere. We discuss downstream applications that should benefit from semantic type information and the semantic embeddings of social entities generated in this work. In particular, we demonstrate enhanced performance on the key task of entity similarity assessment using this information.

In audio and speech processing, tasks usually focus on either the audio or speech modality, even when both sounds and human speech are present in the same audio clip. Recent Auditory Large Language Models (ALLMs) have made it possible to process audio and speech simultaneously within a single model, leading to further considerations of joint audio-speech tasks. In this paper, we investigate how well ALLMs can perform joint audio-speech processing. Specifically, we introduce Joint Audio-Speech Co-Reasoning (JASCO), a novel task that unifies audio and speech processing, strictly requiring co-reasoning across both modalities. We release a scene-reasoning dataset called "What Are They Doing" and establish a joint audio-speech benchmark to evaluate the joint reasoning capability of popular ALLMs. Additionally, we provide deeper insights into the models' behaviors by analyzing their dependence on each modality.

Speech sounds convey a great deal of information about the scenes, resulting in a variety of effects ranging from reverberation to additional ambient sounds. In this paper, we manipulate input speech to sound as though it was recorded within a different scene, given an audio-visual conditional example recorded from that scene. Our model learns through self-supervision, taking advantage of the fact that natural video contains recurring sound events and textures. We extract an audio clip from a video and apply speech enhancement. We then train a latent diffusion model to recover the original speech, using another audio-visual clip taken from elsewhere in the video as a conditional hint. Through this process, the model learns to transfer the conditional example's sound properties to the input speech. We show that our model can be successfully trained using unlabeled, in-the-wild videos, and that an additional visual signal can improve its sound prediction abilities. Please see our project webpage for video results: https://tinglok.netlify.app/files/avsoundscape/

Large language models (LLMs) equipped with chain-of-thoughts (CoT) prompting have shown significant multi-step reasoning capabilities in factual content like mathematics, commonsense, and logic. However, their performance in narrative reasoning, which demands greater abstraction capabilities, remains unexplored. This study utilizes tropes in movie synopses to assess the narrative reasoning abilities of state-of-the-art LLMs and uncovers their low performance. We introduce a trope-wise querying approach to address these challenges and boost the F1 score by 11.8 points. Moreover, Figure 1: While LLMs have revolutionized NLP reasoning, while prior studies suggest that CoT enhances surpassing previous supervised learning (SL) multi-step reasoning, this study shows methods and even reaching human-level performance CoT can cause hallucinations in narrative content, on some tasks, their limitations become apparent when reducing GPT-4's performance. We also tested against the Trope dataset. NLU: Natural Language introduce an Adversarial Injection method to Understanding, CS: Commonsense. Check Section embed trope-related text tokens into movie synopses 1 and 2.2 for details.

While large language models (LLMs) exhibit significant utility across various domains, they simultaneously are susceptible to exploitation for unethical purposes, including academic misconduct and dissemination of misinformation. Consequently, AI-generated text detection systems have emerged as a countermeasure. However, these detection mechanisms demonstrate vulnerability to evasion techniques and lack robustness against textual manipulations. This paper introduces back-translation as a novel technique for evading detection, underscoring the need to enhance the robustness of current detection systems. The proposed method involves translating AI-generated text through multiple languages before back-translating to English. We present a model that combines these back-translated texts to produce a manipulated version of the original AI-generated text. Our findings demonstrate that the manipulated text retains the original semantics while significantly reducing the true positive rate (TPR) of existing detection methods. We evaluate this technique on nine AI detectors, including six open-source and three proprietary systems, revealing their susceptibility to back-translation manipulation. In response to the identified shortcomings of existing AI text detectors, we present a countermeasure to improve the robustness against this form of manipulation. Our results indicate that the TPR of the proposed method declines by only 1.85% after back-translation manipulation. Furthermore, we build a large dataset of 720k texts using eight different LLMs. Our dataset contains both human-authored and LLM-generated texts in various domains and writing styles to assess the performance of our method and existing detectors. This dataset is publicly shared for the benefit of the research community.