NeurIPS 2024 Papers

Solving Olympiad-level mathematical problems represents a significant advancement in machine intelligence and automated reasoning. Current machine learning methods, however, struggle to solve Olympiad-level problems beyond Euclidean plane geometry due to a lack of large-scale, high-quality datasets. The challenge is even greater in algebraic systems, which involve infinite reasoning spaces within finite conditions. To address these issues, we propose *AIPS*, an *Algebraic Inequality Proving System* capable of autonomously generating complex inequality theorems and effectively solving Olympiad-level inequality problems without requiring human demonstrations. During proof search in a mixed reasoning manner, a value curriculum learning strategy on generated datasets is implemented to improve proving performance, demonstrating strong mathematical intuitions. On a test set of 20 International Mathematical Olympiad-level inequality problems, AIPS successfully solved 10, outperforming state-of-the-art methods. Furthermore, AIPS automatically generated a vast array of non-trivial theorems without human intervention, some of which have been evaluated by professional contestants and deemed to reach the level of the International Mathematical Olympiad. Notably, one theorem was selected as a competition problem in a major city's 2024 Mathematical Olympiad. All the materials are available at [sites.google.com/view/aips2](https://sites.google.com/view/aips2)

Compared to classical machine learning (ML) models, generative models offer a new usage paradigm where (i) a single model can be used for many different tasks out-of-the-box; (ii) users interact with this model over a series of natural language prompts; and (iii) the model is ideally evaluated on binary user satisfaction with respect to model outputs. Given these characteristics, we explore the problem of how developers of new generative AI software can release and price their technology. We first develop a comparison of two different models for a specific task with respect to user cost-effectiveness. We then model the pricing problem of generative AI software as a game between two different companies who sequentially release their models before users choose their preferred model for each task. Here, the price optimization problem becomes piecewise continuous where the companies must choose a subset of the tasks on which to be cost-effective and forgo revenue for the remaining tasks. In particular, we reveal the value of market information by showing that a company who deploys later after knowing their competitor’s price can always secure cost-effectiveness on at least one task, whereas the company who is the first-to-market must price their model in a way that incentivizes higher prices from the latecomer in order to gain revenue. Most importantly, we find that if the different tasks are sufficiently similar, the first-to-market model may become cost-ineffective on all tasks regardless of how this technology is priced.

We introduce a novel implicit field representation tailored for multi-person interaction geometry in 3D spaces, capable of simultaneously reconstructing occupancy, instance identification (ID) tags, and contact fields. Volumetric representation of interacting human bodies presents significant challenges, including inaccurately captured geometries, varying degrees of occlusion, and data scarcity. Existing multi-view methods, which either reconstruct each subject in isolation or merge nearby 3D surfaces into a single unified mesh, often fail to capture the intricate geometry between interacting bodies and exploit on datasets with many views and a small group of people for training. Our approach utilizes an implicit representation for interaction geometry contextualized by a multi-view local-global feature module. This module adeptly aggregates both local and global information from individual views and interacting groups, enabling precise modeling of close physical interactions through dense point retrieval in small areas, supported by the implicit fields. Furthermore, we develop a synthetic dataset encompassing diverse multi-person interaction scenarios to enhance the robustness of our geometry estimation. The experimental results demonstrate the superiority of our method to accurately reconstruct human geometries and ID tags within three-dimensional spaces, outperforming conventional multi-view techniques. Notably, our method facilitates unsupervised estimation of contact points without the need for specific training data on contact supervision.

Recent advances in AI models have increased the integration of AI-based decision aids into the human decision making process. To fully unlock the potential of AI-assisted decision making, researchers have computationally modeled how humans incorporate AI recommendations into their final decisions, and utilized these models to improve human-AI team performance. Meanwhile, due to the ``black-box'' nature of AI models, providing AI explanations to human decision makers to help them rely on AI recommendations more appropriately has become a common practice. In this paper, we explore whether we can quantitatively model how humans integrate both AI recommendations and explanations into their decision process, and whether this quantitative understanding of human behavior from the learned model can be utilized to manipulate AI explanations, thereby nudging individuals towards making targeted decisions. Our extensive human experiments across various tasks demonstrate that human behavior can be easily influenced by these manipulated explanations towards targeted outcomes, regardless of the intent being adversarial or benign. Furthermore, individuals often fail to detect any anomalies in these explanations, despite their decisions being affected by them.

Prior research in human-centric AI has primarily addressed single-modality tasks like pedestrian detection, action recognition, and pose estimation. However, the emergence of large multimodal models (LMMs) such as GPT-4V has redirected attention towards integrating language with visual content. Referring expression comprehension (REC) represents a prime example of this multimodal approach. Current human-centric REC benchmarks, typically sourced from general datasets, fall short in the LMM era due to their limitations, such as insufficient testing samples, overly concise referring expressions, and limited vocabulary, making them inadequate for evaluating the full capabilities of modern REC models. In response, we present HC-RefLoCo (Human-Centric Referring Expression Comprehension with Long Context), a benchmark that includes 13,452 images, 24,129 instances, and 44,738 detailed annotations, encompassing a vocabulary of 18,681 words. Each annotation, meticulously reviewed for accuracy, averages 93.2 words and includes topics such as appearance, human-object interaction, location, action, celebrity, and OCR. HC-RefLoCo provides a wider range of instance scales and diverse evaluation protocols, encompassing accuracy with various IoU criteria, scale-aware evaluation, and subject-specific assessments. Our experiments, which assess 24 models, highlight HC-RefLoCo’s potential to advance human-centric AI by challenging contemporary REC models with comprehensive and varied data. Our benchmark, along with the evaluation code, are available at https://github.com/ZhaoJingjing713/HC-RefLoCo.

Gesture synthesis is a vital realm of human-computer interaction, with wide-ranging applications across various fields like film, robotics, and virtual reality. Recent advancements have utilized the diffusion model to improve gesture synthesis. However, the high computational complexity of these techniques limits the application in reality. In this study, we explore the potential of state space models (SSMs). Direct application of SSMs in gesture synthesis encounters difficulties, which stem primarily from the diverse movement dynamics of various body parts. The generated gestures may also exhibit unnatural jittering issues. To address these, we implement a two-stage modeling strategy with discrete motion priors to enhance the quality of gestures. Built upon the selective scan mechanism, we introduce MambaTalk, which integrates hybrid fusion modules, local and global scans to refine latent space representations. Subjective and objective experiments demonstrate that our method surpasses the performance of state-of-the-art models. Our project is publicly available at~\url{https://kkakkkka.github.io/MambaTalk/}.

Large-scale generative language and vision-language models (LLMs and VLMs) excel in few-shot in-context learning for decision making and instruction following. However, they require high-quality exemplar demonstrations to be included in their context window. In this work, we ask: Can LLMs and VLMs generate their own examples from generic, sub-optimal demonstrations? We propose In-Context Abstraction Learning (ICAL), a method that builds a memory of multimodal experience from sub-optimal demonstrations and human feedback. Given a task demonstration that may contain inefficiencies or mistakes, a VLM abstracts the trajectory into a generalized program by correcting inefficient actions and annotating cognitive abstractions: causal relationships, object state changes, temporal subgoals, and task-relevant visual elements. These abstractions are iteratively improved and adapted through human feedback while the agent attempts to execute the trajectory in a similar environment. The resulting examples, when used as exemplars in the prompt, significantly improve decision-making in retrieval-augmented LLM and VLM agents. Moreover, as the agent's library of examples grows, it becomes more efficient, relying less on human feedback and requiring fewer environment interactions per demonstration. Our ICAL agent surpasses the state-of-the-art in dialogue-based instruction following in TEACh, multimodal web agents in VisualWebArena, and action anticipation in Ego4D. In TEACh, we achieve a 12.6% improvement in goal-condition success. In VisualWebArena, our task success rate improves over the SOTA from 14.3% to 22.7% using GPT4V. In Ego4D action forecasting, we improve over few-shot GPT-4V and remain competitive with supervised models. We show finetuning our retrieval-augmented in-context agent yields additional improvements. Our approach significantly reduces reliance on manual prompt engineering and consistently outperforms in-context learning from action plans that lack such abstractions.