The Turing test examines whether AIs can exhibit human-like behaviour in natural language conversations. Traditional Turing tests adopt a rigid dialogue format where each participant sends only one message each time and require continuous human involvement to direct the entire interaction with the test subject. This fails to reflect a natural conversational style and hinders the evaluation of Large Language Models (LLMs) in complex and prolonged dialogues. This paper proposes the Self-Directed Turing Test, which extends the original test with a burst dialogue format, allowing more dynamic exchanges by multiple consecutive messages. It further efficiently reduces human workload by having the LLM self-direct the majority of the test process, iteratively generating dialogues that simulate its interaction with humans. With the pseudo-dialogue history, the model then engages in a shorter dialogue with a human, which is paired with a human-human conversation on the same topic to be judged using questionnaires. We introduce the X-Turn Pass-Rate metric to assess the human likeness of LLMs across varying durations. While LLMs like GPT-4 initially perform well, achieving pass rates of 51.9% and 38.9% during 3 turns and 10 turns of dialogues respectively, their performance drops as the dialogue progresses, which underscores the difficulty in maintaining consistency in the long term.

Suppose Alice has a distribution $P$ and Bob has a distribution $Q$. Alice wants to generate a sample $a\sim P$ and Bob a sample $b \sim Q$ such that $a = b$ with has as high of probability as possible. It is well-known that, by sampling from an optimal coupling between the distributions, Alice and Bob can achieve $Pr[a = b] = 1 - D_{TV}(P,Q)$, where $D_{TV}(P,Q)$ is the total variation distance. What if Alice and Bob must solve this same problem without communicating at all? Perhaps surprisingly, with access to public randomness, they can still achieve $Pr[a=b] \geq \frac{1-D_{TV}(P,Q)}{1+D_{TV}(P,Q)} \geq 1-2D_{TV}(P,Q)$. In fact, this bound can be obtained using a simple protocol based on the Weighted MinHash algorithm. In this work, we explore the communication-free coupling problem in greater depth. First, we show that an equally simple protocol based on Gumbel sampling matches the worst-case guarantees of the Weighted MinHash approach, but tends to perform better in practice. Conversely, we prove that both approaches are actually sharp: no communication-free protocol can achieve $Pr[a=b]>\frac{1-D_{TV}(P,Q)}{1+D_{TV}(P,Q)}$ in the worst-case. Finally, we prove that, for distributions over $n$ items, there exists a scheme that uses just $O(\log(n/\epsilon))$ bits of communication to achieve $Pr[a = b] = 1 - D_{TV}(P,Q) - \epsilon$, i.e. to essentially match optimal coupling. Beyond our theoretical results, we demonstrate an application of communication-free coupling to speculative decoding, a recent method for accelerating autoregressive large language models [Leviathan, Kalman, Matias, ICML 2023]. We show that communication-free protocols yield a variant of speculative decoding that we call Drafter-Invariant Speculative Decoding, which has the desirable property that the output of the method is fixed given a fixed random seed, regardless of what drafter is used for speculation.

Chip design relies heavily on generating Boolean circuits, such as AND-Inverter Graphs (AIGs), from functional descriptions like truth tables. While recent advances in deep learning have aimed to accelerate circuit design, these efforts have mostly focused on tasks other than synthesis, and traditional heuristic methods have plateaued. In this paper, we introduce ShortCircuit, a novel transformerbased architecture that leverages the structural properties of AIGs and performs efficient space exploration. Contrary to prior approaches attempting end-to-end generation of logic circuits using deep networks, ShortCircuit employs a two-phase process combining supervised with reinforcement learning to enhance generalization to unseen truth tables. We also propose an AlphaZero variant to handle the double exponentially large state space and the sparsity of the rewards, enabling the discovery of near-optimal designs. To evaluate the generative performance of our trained model, we extract 500 truth tables from a benchmark set of 20 realworld circuits. ShortCircuit successfully generates AIGs for84.6% of the 8-input test truth tables, and outperforms the state-of-the-art logic synthesis tool, ABC, by 14.61% in terms of circuits size. The rapid proliferation of AI has triggered an unprecedented surge in computational demands, exceeding the capabilities of existing hardware and thereby becoming a major bottleneck to AI's continued growth. To enable further advancements in AI and other domains, the development of new hardware is crucial. Chip design plays a pivotal role in this development effort allowing the nextgeneration of computing systems to emerge. However, traditional methodologies struggle to keep pace with the accelerating demands, underscoring the need for innovative approaches to accelerate the design process and discover novel architectures. At its core, a chip is the physical embodiment of a Boolean function, transforming binary inputs into desired outputs.

Although instruction tuning is widely used to adjust behavior in Large Language Models (LLMs), extensive empirical evidence and research indicates that it is primarily a process where the model fits to specific task formats, rather than acquiring new knowledge or capabilities. We propose that this limitation stems from biased features learned during instruction tuning, which differ from ideal task-specfic features, leading to learn less underlying semantics in downstream tasks. However, ideal features are unknown and incalculable, constraining past work to rely on prior knowledge to assist reasoning or training, which limits LLMs' capabilities to the developers' abilities, rather than data-driven scalable learning. In our paper, through our novel data synthesis method, DELIA (Diversity-Enhanced Learning for Instruction Adaptation), we leverage the buffering effect of extensive diverse data in LLMs training to transform biased features in instruction tuning into approximations of ideal features, without explicit prior ideal features. Experiments show DELIA's better performance compared to common instruction tuning and other baselines. It outperforms common instruction tuning by 17.07%-33.41% on Icelandic-English translation bleurt score (WMT-21 dataset, gemma-7b-it) and improves accuracy by 36.1% on formatted text generation (Llama2-7b-chat). Notably, among knowledge injection methods we've known, DELIA uniquely align the internal representations of new special tokens with their prior semantics.

Large semantic knowledge bases are grounded in factual knowledge. However, recent approaches to dense text representations (embeddings) do not efficiently exploit these resources. Dense and robust representations of documents are essential for effectively solving downstream classification and retrieval tasks. This work demonstrates that injecting embedded information from knowledge bases can augment the performance of contemporary Large Language Model (LLM)-based representations for the task of text classification. Further, by considering automated machine learning (AutoML) with the fused representation space, we demonstrate it is possible to improve classification accuracy even if we use low-dimensional projections of the original representation space obtained via efficient matrix factorization. This result shows that significantly faster classifiers can be achieved with minimal or no loss in predictive performance, as demonstrated using five strong LLM baselines on six diverse real-life datasets.

Recent studies have begun developing autonomous game players for social deduction games using large language models (LLMs). When building LLM players, fine-grained evaluations are crucial for addressing weaknesses in game-playing abilities. However, existing studies have often overlooked such assessments. Specifically, we point out two issues with the evaluation methods employed. First, game-playing abilities have typically been assessed through game-level outcomes rather than specific event-level skills; Second, error analyses have lacked structured methodologies. To address these issues, we propose an approach utilizing a variant of the SpyFall game, named SpyGame. We conducted an experiment with four LLMs, analyzing their gameplay behavior in SpyGame both quantitatively and qualitatively. For the quantitative analysis, we introduced eight metrics to resolve the first issue, revealing that these metrics are more effective than existing ones for evaluating the two critical skills: intent identification and camouflage. In the qualitative analysis, we performed thematic analysis to resolve the second issue. This analysis identifies four major categories that affect gameplay of LLMs. Additionally, we demonstrate how these categories complement and support the findings from the quantitative analysis.

Federated learning (FL) has shown promising results on training machine learning models with privacy preservation. However, for large models with over 100 million parameters, the training resource requirement becomes an obstacle for FL because common devices do not have enough memory and computation power to finish the FL tasks. Although efficient training methods have been proposed, it is still a challenge to train the large models like Conformer based ASR. This paper presents a systematic solution to train the full-size ASR models of 130M parameters with FL. To our knowledge, this is the first real-world FL application of the Conformer model, which is also the largest model ever trained with FL so far. And this is the first paper showing FL can improve the ASR model quality with a set of proposed methods to refine the quality of data and labels of clients. We demonstrate both the training efficiency and the model quality improvement in real-world experiments.

Pedestrian Attribute Recognition (PAR) is one of the indispensable tasks in human-centered research. However, existing datasets neglect different domains (e.g., environments, times, populations, and data sources), only conducting simple random splits, and the performance of these datasets has already approached saturation. In the past five years, no large-scale dataset has been opened to the public. To address this issue, this paper proposes a new large-scale, cross-domain pedestrian attribute recognition dataset to fill the data gap, termed MSP60K. It consists of 60,122 images and 57 attribute annotations across eight scenarios. Synthetic degradation is also conducted to further narrow the gap between the dataset and real-world challenging scenarios. To establish a more rigorous benchmark, we evaluate 17 representative PAR models under both random and cross-domain split protocols on our dataset. Additionally, we propose an innovative Large Language Model (LLM) augmented PAR framework, named LLM-PAR. This framework processes pedestrian images through a Vision Transformer (ViT) backbone to extract features and introduces a multi-embedding query Transformer to learn partial-aware features for attribute classification. Significantly, we enhance this framework with LLM for ensemble learning and visual feature augmentation. Comprehensive experiments across multiple PAR benchmark datasets have thoroughly validated the efficacy of our proposed framework. The dataset and source code accompanying this paper will be made publicly available at \url{https://github.com/Event-AHU/OpenPAR}.

Mobile devices, especially smartphones, can support rich functions and have developed into indispensable tools in daily life. With the rise of generative AI services, smartphones can potentially transform into personalized assistants, anticipating user needs and scheduling services accordingly. Predicting user intents on smartphones, and reflecting anticipated activities based on past interactions and context, remains a pivotal step towards this vision. Existing research predominantly focuses on specific domains, neglecting the challenge of modeling diverse event sequences across dynamic contexts. Leveraging pre-trained language models (PLMs) offers a promising avenue, yet adapting PLMs to on-device user intent prediction presents significant challenges. To address these challenges, we propose PITuning, a Population-to-Individual Tuning framework. PITuning enhances common pattern extraction through dynamic event-to-intent transition modeling and addresses long-tailed preferences via adaptive unlearning strategies. Experimental results on real-world datasets demonstrate PITuning's superior intent prediction performance, highlighting its ability to capture long-tailed preferences and its practicality for on-device prediction scenarios.

Large language models (LLMs) have shown remarkable capability in numerous tasks and applications. However, fine-tuning LLMs using high-quality datasets under external supervision remains prohibitively expensive. In response, LLM self-improvement approaches have been vibrantly developed recently. The typical paradigm of LLM self-improvement involves training LLM on self-generated data, part of which may be detrimental and should be filtered out due to the unstable data quality. While current works primarily employs filtering strategies based on answer correctness, in this paper, we demonstrate that filtering out correct but with high distribution shift extent (DSE) samples could also benefit the results of self-improvement. Given that the actual sample distribution is usually inaccessible, we propose a new metric called DS weight to approximate DSE, inspired by the Importance Weighting methods. Consequently, we integrate DS weight with self-consistency to comprehensively filter the self-generated samples and fine-tune the language model. Experiments show that with only a tiny valid set (up to 5\% size of the training set) to compute DS weight, our approach can notably promote the reasoning ability of current LLM self-improvement methods. The resulting performance is on par with methods that rely on external supervision from pre-trained reward models.