Daniel Zhang

Daniel Zhang, Nestor Maslej, Erik Brynjolfsson et al. · salesforce, stanford

Welcome to the fifth edition of the AI Index Report! The latest edition includes data from a broad set of academic, private, and nonprofit organizations as well as more self-collected data and original analysis than any previous editions, including an expanded technical performance chapter, a new survey of robotics researchers around the world, data on global AI legislation records in 25 countries, and a new chapter with an in-depth analysis of technical AI ethics metrics. The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Its mission is to provide unbiased, rigorously vetted, and globally sourced data for policymakers, researchers, executives, journalists, and the general public to develop a more thorough and nuanced understanding of the complex field of AI. The report aims to be the world's most credible and authoritative source for data and insights about AI.

Amazon AGI, Aaron Langford, Aayush Shah et al. · amazon-science

We present Amazon Nova, a new generation of state-of-the-art foundation models that deliver frontier intelligence and industry-leading price performance. Amazon Nova Pro is a highly-capable multimodal model with the best combination of accuracy, speed, and cost for a wide range of tasks. Amazon Nova Lite is a low-cost multimodal model that is lightning fast for processing images, video, documents and text. Amazon Nova Micro is a text-only model that delivers our lowest-latency responses at very low cost. Amazon Nova Canvas is an image generation model that creates professional grade images with rich customization controls. Amazon Nova Reel is a video generation model offering high-quality outputs, customization, and motion control. Our models were built responsibly and with a commitment to customer trust, security, and reliability. We report benchmarking results for core capabilities, agentic performance, long context, functional adaptation, runtime performance, and human evaluation.

Zain Shamsi, Daniel Zhang, Daehyun Kyoung et al.

Network honeypots are often used by information security teams to measure the threat landscape in order to secure their networks. With the advancement of honeypot development, today's medium-interaction honeypots provide a way for security teams and researchers to deploy these active defense tools that require little maintenance on a variety of protocols. In this work, we deploy such honeypots on five different protocols on the public Internet and study the intent and sophistication of the attacks we observe. We then use the information gained to develop a clustering approach that identifies correlations in attacker behavior to discover IPs that are highly likely to be controlled by a single operator, illustrating the advantage of using these honeypots for data collection.

M Charity, Yash Bhartia, Daniel Zhang et al.

This paper introduces a system used to generate game feature suggestions based on a text prompt. Trained on the game descriptions of almost 60k games, it uses the word embeddings of a small GLoVe model to extract features and entities found in thematically similar games which are then passed through a generator model to generate new features for a user's prompt. We perform a short user study comparing the features generated from a fine-tuned GPT-2 model, a model using the ConceptNet, and human-authored game features. Although human suggestions won the overall majority of votes, the GPT-2 model outperformed the human suggestions in certain games. This system is part of a larger game design assistant tool that is able to collaborate with users at a conceptual level.

Rishi Bommasani, Kevin Klyman, Shayne Longpre et al.

Foundation models have rapidly permeated society, catalyzing a wave of generative AI applications spanning enterprise and consumer-facing contexts. While the societal impact of foundation models is growing, transparency is on the decline, mirroring the opacity that has plagued past digital technologies (e.g. social media). Reversing this trend is essential: transparency is a vital precondition for public accountability, scientific innovation, and effective governance. To assess the transparency of the foundation model ecosystem and help improve transparency over time, we introduce the Foundation Model Transparency Index. The Foundation Model Transparency Index specifies 100 fine-grained indicators that comprehensively codify transparency for foundation models, spanning the upstream resources used to build a foundation model (e.g data, labor, compute), details about the model itself (e.g. size, capabilities, risks), and the downstream use (e.g. distribution channels, usage policies, affected geographies). We score 10 major foundation model developers (e.g. OpenAI, Google, Meta) against the 100 indicators to assess their transparency. To facilitate and standardize assessment, we score developers in relation to their practices for their flagship foundation model (e.g. GPT-4 for OpenAI, PaLM 2 for Google, Llama 2 for Meta). We present 10 top-level findings about the foundation model ecosystem: for example, no developer currently discloses significant information about the downstream impact of its flagship model, such as the number of users, affected market sectors, or how users can seek redress for harm. Overall, the Foundation Model Transparency Index establishes the level of transparency today to drive progress on foundation model governance via industry standards and regulatory intervention.

Johannes Köhler, Daniel Zhang, Raffaele Soloperto et al.

We present a model predictive control (MPC) framework for efficient navigation of mobile robots in cluttered environments. The proposed approach integrates a finite-segment shortest path planner into the finite-horizon trajectory optimization of the MPC. This formulation ensures convergence to dynamically selected targets and guarantees collision avoidance, even under general nonlinear dynamics and cluttered environments. The approach is validated through hardware experiments on a small ground robot, where a human operator dynamically assigns target locations that a robot should reach while avoiding obstacles. The robot reached new targets within 2-3 seconds and responded to new commands within 50 ms to 100 ms, immediately adjusting its motion even while still moving at high speeds toward a previous target.

Ali Abbasi, Chayne Thrash, Elaheh Akbari et al.

The rapid progress of AI, combined with its unprecedented public adoption and the propensity of large neural networks to memorize training data, has given rise to significant data privacy concerns. To address these concerns, machine unlearning has emerged as an essential technique to selectively remove the influence of specific training data points on trained models. In this paper, we approach the machine unlearning problem through the lens of continual learning. Given a trained model and a subset of training data designated to be forgotten (i.e., the "forget set"), we introduce a three-step process, named CovarNav, to facilitate this forgetting. Firstly, we derive a proxy for the model's training data using a model inversion attack. Secondly, we mislabel the forget set by selecting the most probable class that deviates from the actual ground truth. Lastly, we deploy a gradient projection method to minimize the cross-entropy loss on the modified forget set (i.e., learn incorrect labels for this set) while preventing forgetting of the inverted samples. We rigorously evaluate CovarNav on the CIFAR-10 and Vggface2 datasets, comparing our results with recent benchmarks in the field and demonstrating the efficacy of our proposed approach.

Gil Halevi, Daniel Zhang, Jason Zhang

Given two sets of points A and B, $|A| = m$, $|B| = n$, the Chamfer distance from $A$ to $B$ is defined as $\operatorname{CD}(A,B) = \sum_{a\in A} \min_{b\in B} d(a,b)$, where $d$ is a distance metric. Chamfer distance is a popular measure of dissimilarity between two sets of points that has seen increasing usage in computer vision and information retrieval as a substitute for the more computationally demanding Earth Mover's distance. We propose a new problem, Chamfer distance under translation, defined as $\operatorname{CDuT}(A,B) :=\min_{t\in \mathbb{R}^d} \operatorname{CD}(A+t,B)$, where $A+t$ denotes the translation of every point in $A$ by $t$. Chamfer distance under translation is valuable in cases where translations capture aspects of the data unlikely to be relevant for dissimilarity, such as temporal, spatial, or other semantic information. For Chamfer distance under translation, we provide four algorithms: (1) an exact quadratic time algorithm in one dimension, (2) a near quadratic time ($2+\varepsilon$)-approximation algorithm in higher dimensions, (3) a $(1+\varepsilon)$-approximation algorithm with running time $\mathcal{O}(mn^2\varepsilon^{-(d+1)})$, and (4) a near-quadratic time $(1+\varepsilon)$-approximation algorithm for answering the decision version of $\operatorname{CDuT}$ given a separation assumption on $B$. We additionally explore the fine-grained complexity of $\operatorname{CDuT}$.

Daniel Zhang, Vikram Voleti, Alexander Wong et al.

Creating high-performance generalizable deep neural networks for phytoplankton monitoring requires utilizing large-scale data coming from diverse global water sources. A major challenge to training such networks lies in data privacy, where data collected at different facilities are often restricted from being transferred to a centralized location. A promising approach to overcome this challenge is federated learning, where training is done at site level on local data, and only the model parameters are exchanged over the network to generate a global model. In this study, we explore the feasibility of leveraging federated learning for privacy-preserving training of deep neural networks for phytoplankton classification. More specifically, we simulate two different federated learning frameworks, federated learning (FL) and mutually exclusive FL (ME-FL), and compare their performance to a traditional centralized learning (CL) framework. Experimental results from this study demonstrate the feasibility and potential of federated learning for phytoplankton monitoring.

Judy Hanwen Shen, Ken Liu, Angelina Wang et al.

Data transparency has emerged as a rallying cry for addressing concerns about AI: data quality, privacy, and copyright chief among them. Yet while these calls are crucial for accountability, current transparency policies often fall short of their intended aims. Similar to nutrition facts for food, policies aimed at nutrition facts for AI currently suffer from a limited consideration of research on effective disclosures. We offer an institutional perspective and identify three common fallacies in policy implementations of data disclosures for AI. First, many data transparency proposals exhibit a specification gap between the stated goals of data transparency and the actual disclosures necessary to achieve such goals. Second, reform attempts exhibit an enforcement gap between required disclosures on paper and enforcement to ensure compliance in fact. Third, policy proposals manifest an impact gap between disclosed information and meaningful changes in developer practices and public understanding. Informed by the social science on transparency, our analysis identifies affirmative paths for transparency that are effective rather than merely symbolic.

Daniel Zhang

Time series data are observations collected over time intervals. Successful analysis of time series data captures patterns such as trends, cyclicity and irregularity, which are crucial for decision making in research, business, and governance. However, missing values in time series data occur often and present obstacles to successful analysis, thus they need to be filled with alternative values, a process called imputation. Although various approaches have been attempted for robust imputation of time series data, even the most advanced methods still face challenges including limited scalability, poor capacity to handle heterogeneous data types and inflexibility due to requiring strong assumptions of data missing mechanisms. Moreover, the imputation accuracy of these methods still has room for improvement. In this study, I developed tsDataWig (time-series DataWig) by modifying DataWig, a neural network-based method that possesses the capacity to process large datasets and heterogeneous data types but was designed for non-time series data imputation. Unlike the original DataWig, tsDataWig can directly handle values of time variables and impute missing values in complex time series datasets. Using one simulated and three different complex real-world time series datasets, I demonstrated that tsDataWig outperforms the original DataWig and the current state-of-the-art methods for time series data imputation and potentially has broad application due to not requiring strong assumptions of data missing mechanisms. This study provides a valuable solution for robustly imputing missing values in challenging time series datasets, which often contain millions of samples, high dimensional variables, and heterogeneous data types.

Sayash Kapoor, Rishi Bommasani, Kevin Klyman et al.

Foundation models are powerful technologies: how they are released publicly directly shapes their societal impact. In this position paper, we focus on open foundation models, defined here as those with broadly available model weights (e.g. Llama 2, Stable Diffusion XL). We identify five distinctive properties (e.g. greater customizability, poor monitoring) of open foundation models that lead to both their benefits and risks. Open foundation models present significant benefits, with some caveats, that span innovation, competition, the distribution of decision-making power, and transparency. To understand their risks of misuse, we design a risk assessment framework for analyzing their marginal risk. Across several misuse vectors (e.g. cyberattacks, bioweapons), we find that current research is insufficient to effectively characterize the marginal risk of open foundation models relative to pre-existing technologies. The framework helps explain why the marginal risk is low in some cases, clarifies disagreements about misuse risks by revealing that past work has focused on different subsets of the framework with different assumptions, and articulates a way forward for more constructive debate. Overall, our work helps support a more grounded assessment of the societal impact of open foundation models by outlining what research is needed to empirically validate their theoretical benefits and risks.

Austin Jia, Avaneesh Ramesh, Zain Shamsi et al.

Retrieval-Augmented Generation (RAG) has emerged as the dominant architectural pattern to operationalize Large Language Model (LLM) usage in Cyber Threat Intelligence (CTI) systems. However, this design is susceptible to poisoning attacks, and previously proposed defenses can fail for CTI contexts as cyber threat information is often completely new for emerging attacks, and sophisticated threat actors can mimic legitimate formats, terminology, and stylistic conventions. To address this issue, we propose that the robustness of modern RAG defenses can be accelerated by applying source credibility algorithms on corpora, using PageRank as an example. In our experiments, we demonstrate quantitatively that our algorithm applies a lower authority score to malicious documents while promoting trusted content, using the standardized MS MARCO dataset. We also demonstrate proof-of-concept performance of our algorithm on CTI documents and feeds.

Daniel Son, Sanjana Rathore, Andrew Rufail et al.

We investigate feature universality in Gemma-2 language models (Gemma-2-2B and Gemma-2-9B), asking whether models with a four-fold difference in scale still converge on comparable internal concepts. Using the Sparse Autoencoder (SAE) dictionary-learning pipeline, we utilize SAEs on each model's residual-stream activations, align the resulting monosemantic features via activation correlation, and compare the matched feature spaces with SVCCA and RSA. Middle layers yield the strongest overlap, while early and late layers show far less similarity. Preliminary experiments extend the analysis from single tokens to multi-token subspaces, showing that semantically similar subspaces interact similarly with language models. These results strengthen the case that large language models carve the world into broadly similar, interpretable features despite size differences, reinforcing universality as a foundation for cross-model interpretability.

Daniel Zhang, Saurabh Mishra, Erik Brynjolfsson et al.

Welcome to the fourth edition of the AI Index Report. This year we significantly expanded the amount of data available in the report, worked with a broader set of external organizations to calibrate our data, and deepened our connections with the Stanford Institute for Human-Centered Artificial Intelligence (HAI). The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Its mission is to provide unbiased, rigorously vetted, and globally sourced data for policymakers, researchers, executives, journalists, and the general public to develop intuitions about the complex field of AI. The report aims to be the most credible and authoritative source for data and insights about AI in the world.

Daniel Zhang, Colleen P. Bailey

In this paper, we investigate the obstacle avoidance and navigation problem in the robotic control area. For solving such a problem, we propose revised Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization algorithms with an improved reward shaping technique. We compare the performances between the original DDPG and PPO with the revised version of both on simulations with a real mobile robot and demonstrate that the proposed algorithms achieve better results.

Yang Zhang, Daniel Zhang, Nathan Vance et al.

Social sensing has emerged as a new sensing paradigm where humans (or devices on their behalf) collectively report measurements about the physical world. This paper focuses on a quality-cost-aware task allocation problem in multi-attribute social sensing applications. The goal is to identify a task allocation strategy (i.e., decide when and where to collect sensing data) to achieve an optimized tradeoff between the data quality and the sensing cost. While recent progress has been made to tackle similar problems, three important challenges have not been well addressed: (i) "online task allocation": the task allocation schemes need to respond quickly to the potentially large dynamics of the measured variables in social sensing; (ii) "multi-attribute constrained optimization": minimizing the overall sensing error given the dependencies and constraints of multiple attributes of the measured variables is a non-trivial problem to solve; (iii) "nonuniform task allocation cost": the task allocation cost in social sensing often has a nonuniform distribution which adds additional complexity to the optimized task allocation problem. This paper develops a Quality-Cost-Aware Online Task Allocation (QCO-TA) scheme to address the above challenges using a principled online reinforcement learning framework. We evaluate the QCO-TA scheme through a real-world social sensing application and the results show that our scheme significantly outperforms the state-of-the-art baselines in terms of both sensing accuracy and cost.

Lanyu Shang, Daniel Zhang, Michael Wang et al.

Online video sharing platforms (e.g., YouTube, Vimeo) have become an increasingly popular paradigm for people to consume video contents. Clickbait video, whose content clearly deviates from its title/thumbnail, has emerged as a critical problem on online video sharing platforms. Current clickbait detection solutions that mainly focus on analyzing the text of the title, the image of the thumbnail, or the content of the video are shown to be suboptimal in detecting the online clickbait videos. In this paper, we develop a novel content-agnostic scheme, Online Video Clickbait Protector (OVCP), to effectively detect clickbait videos by exploring the comments from the audience who watched the video. Different from existing solutions, OVCP does not directly analyze the content of the video and its pre-click information (e.g., title and thumbnail). Therefore, it is robust against sophisticated content creators who often generate clickbait videos that can bypass the current clickbait detectors. We evaluate OVCP with a real-world dataset collected from YouTube. Experimental results demonstrate that OVCP is effective in identifying clickbait videos and significantly outperforms both state-of-the-art baseline models and human annotators.

Daniel Zhang, Nathan Vance, Dong Wang

This paper overviews the state of the art, research challenges, and future opportunities in an emerging research direction: Social Sensing based Edge Computing (SSEC). Social sensing has emerged as a new sensing application paradigm where measurements about the physical world are collected from humans or from devices on their behalf. The advent of edge computing pushes the frontier of computation, service, and data along the cloud-to-things continuum. The merging of these two technical trends generates a set of new research challenges that need to be addressed. In this paper, we first define the new SSEC paradigm that is motivated by a few underlying technology trends. We then present a few representative real-world case studies of SSEC applications and several key research challenges that exist in those applications. Finally, we envision a few exciting research directions in future SSEC. We hope this paper will stimulate discussions of this emerging research direction in the community.

Jacob Abernethy, Cyrus Anderson, Alex Chojnacki et al.

Performing arts organizations aim to enrich their communities through the arts. To do this, they strive to match their performance offerings to the taste of those communities. Success relies on understanding audience preference and predicting their behavior. Similar to most e-commerce or digital entertainment firms, arts presenters need to recommend the right performance to the right customer at the right time. As part of the Michigan Data Science Team (MDST), we partnered with the University Musical Society (UMS), a non-profit performing arts presenter housed in the University of Michigan, Ann Arbor. We are providing UMS with analysis and business intelligence, utilizing historical individual-level sales data. We built a recommendation system based on collaborative filtering, gaining insights into the artistic preferences of customers, along with the similarities between performances. To better understand audience behavior, we used statistical methods from customer-base analysis. We characterized customer heterogeneity via segmentation, and we modeled customer cohorts to understand and predict ticket purchasing patterns. Finally, we combined statistical modeling with natural language processing (NLP) to explore the impact of wording in program descriptions. These ongoing efforts provide a platform to launch targeted marketing campaigns, helping UMS carry out its mission by allocating its resources more efficiently. Celebrating its 138th season, UMS is a 2014 recipient of the National Medal of Arts, and it continues to enrich communities by connecting world-renowned artists with diverse audiences, especially students in their formative years. We aim to contribute to that mission through data science and customer analytics.