arXiv subjects

Arvind Narayanan

Publications and source records attributed to Arvind Narayanan.

At least 37 records · Page 2Linked to original sources

arXiv2025

The Responsible Foundation Model Development Cheatsheet: A Review of Tools & Resources

Foundation model development attracts a rapidly expanding body of contributors, scientists, and applications. To help shape responsible development practices, we introduce the Foundation Model Development Cheatsheet: a growing collection of 250+ tools and resources spanning text, vision, and speech modalities. We draw on a large body of prior work to survey resources (e.g. software, documentation, frameworks, guides, and practical tools) that support informed data selection, processing, and understanding, precise and limitation-aware artifact documentation, efficient model training, advance awareness of the environmental impact from training, careful model evaluation of capabilities, risks, and claims, as well as responsible model release, licensing and deployment practices. We hope this curated collection of resources helps guide more responsible development. The process of curating this list, enabled us to review the AI development ecosystem, revealing what tools are critically missing, misused, or over-used in existing practices. We find that (i) tools for data sourcing, model evaluation, and monitoring are critically under-serving ethical and real-world needs, (ii) evaluations for model safety, capabilities, and environmental impact all lack reproducibility and transparency, (iii) text and particularly English-centric analyses continue to dominate over multilingual and multi-modal analyses, and (iv) evaluation of systems, rather than just models, is needed so that capabilities and impact are assessed in context.

cs.LG↗

arXiv2025

International AI Safety Report

The first International AI Safety Report comprehensively synthesizes the current evidence on the capabilities, risks, and safety of advanced AI systems. The report was mandated by the nations attending the AI Safety Summit in Bletchley, UK. Thirty nations, the UN, the OECD, and the EU each nominated a representative to the report's Expert Advisory Panel. A total of 100 AI experts contributed, representing diverse perspectives and disciplines. Led by the report's Chair, these independent experts collectively had full discretion over the report's content.

cs.CY↗

arXiv2024

AI Agents That Matter

AI agents are an exciting new research direction, and agent development is driven by benchmarks. Our analysis of current agent benchmarks and evaluation practices reveals several shortcomings that hinder their usefulness in real-world applications. First, there is a narrow focus on accuracy without attention to other metrics. As a result, SOTA agents are needlessly complex and costly, and the community has reached mistaken conclusions about the sources of accuracy gains. Our focus on cost in addition to accuracy motivates the new goal of jointly optimizing the two metrics. We design and implement one such optimization, showing its potential to greatly reduce cost while maintaining accuracy. Second, the benchmarking needs of model and downstream developers have been conflated, making it hard to identify which agent would be best suited for a particular application. Third, many agent benchmarks have inadequate holdout sets, and sometimes none at all. This has led to agents that are fragile because they take shortcuts and overfit to the benchmark in various ways. We prescribe a principled framework for avoiding overfitting. Finally, there is a lack of standardization in evaluation practices, leading to a pervasive lack of reproducibility. We hope that the steps we introduce for addressing these shortcomings will spur the development of agents that are useful in the real world and not just accurate on benchmarks.

cs.LG↗

arXiv2024

AI Risk Management Should Incorporate Both Safety and Security

The exposure of security vulnerabilities in safety-aligned language models, e.g., susceptibility to adversarial attacks, has shed light on the intricate interplay between AI safety and AI security. Although the two disciplines now come together under the overarching goal of AI risk management, they have historically evolved separately, giving rise to differing perspectives. Therefore, in this paper, we advocate that stakeholders in AI risk management should be aware of the nuances, synergies, and interplay between safety and security, and unambiguously take into account the perspectives of both disciplines in order to devise mostly effective and holistic risk mitigation approaches. Unfortunately, this vision is often obfuscated, as the definitions of the basic concepts of "safety" and "security" themselves are often inconsistent and lack consensus across communities. With AI risk management being increasingly cross-disciplinary, this issue is particularly salient. In light of this conceptual challenge, we introduce a unified reference framework to clarify the differences and interplay between AI safety and AI security, aiming to facilitate a shared understanding and effective collaboration across communities.

cs.CR↗

arXiv2024

A Safe Harbor for AI Evaluation and Red Teaming

Independent evaluation and red teaming are critical for identifying the risks posed by generative AI systems. However, the terms of service and enforcement strategies used by prominent AI companies to deter model misuse have disincentives on good faith safety evaluations. This causes some researchers to fear that conducting such research or releasing their findings will result in account suspensions or legal reprisal. Although some companies offer researcher access programs, they are an inadequate substitute for independent research access, as they have limited community representation, receive inadequate funding, and lack independence from corporate incentives. We propose that major AI developers commit to providing a legal and technical safe harbor, indemnifying public interest safety research and protecting it from the threat of account suspensions or legal reprisal. These proposals emerged from our collective experience conducting safety, privacy, and trustworthiness research on generative AI systems, where norms and incentives could be better aligned with public interests, without exacerbating model misuse. We believe these commitments are a necessary step towards more inclusive and unimpeded community efforts to tackle the risks of generative AI.

cs.AI↗

arXiv2024

On the Societal Impact of Open Foundation Models

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.

cs.CY↗

arXiv2024

Foundation Model Transparency Reports

Foundation models are critical digital technologies with sweeping societal impact that necessitates transparency. To codify how foundation model developers should provide transparency about the development and deployment of their models, we propose Foundation Model Transparency Reports, drawing upon the transparency reporting practices in social media. While external documentation of societal harms prompted social media transparency reports, our objective is to institutionalize transparency reporting for foundation models while the industry is still nascent. To design our reports, we identify 6 design principles given the successes and shortcomings of social media transparency reporting. To further schematize our reports, we draw upon the 100 transparency indicators from the Foundation Model Transparency Index. Given these indicators, we measure the extent to which they overlap with the transparency requirements included in six prominent government policies (e.g., the EU AI Act, the US Executive Order on Safe, Secure, and Trustworthy AI). Well-designed transparency reports could reduce compliance costs, in part due to overlapping regulatory requirements across different jurisdictions. We encourage foundation model developers to regularly publish transparency reports, building upon recommendations from the G7 and the White House.

cs.LG↗

arXiv2024

Promises and pitfalls of artificial intelligence for legal applications

Is AI set to redefine the legal profession? We argue that this claim is not supported by the current evidence. We dive into AI's increasingly prevalent roles in three types of legal tasks: information processing; tasks involving creativity, reasoning, or judgment; and predictions about the future. We find that the ease of evaluating legal applications varies greatly across legal tasks, based on the ease of identifying correct answers and the observability of information relevant to the task at hand. Tasks that would lead to the most significant changes to the legal profession are also the ones most prone to overoptimism about AI capabilities, as they are harder to evaluate. We make recommendations for better evaluation and deployment of AI in legal contexts.

cs.CY↗

arXiv2023

REFORMS: Reporting Standards for Machine Learning Based Science

Machine learning (ML) methods are proliferating in scientific research. However, the adoption of these methods has been accompanied by failures of validity, reproducibility, and generalizability. These failures can hinder scientific progress, lead to false consensus around invalid claims, and undermine the credibility of ML-based science. ML methods are often applied and fail in similar ways across disciplines. Motivated by this observation, our goal is to provide clear reporting standards for ML-based science. Drawing from an extensive review of past literature, we present the REFORMS checklist ($\textbf{Re}$porting Standards $\textbf{For}$ $\textbf{M}$achine Learning Based $\textbf{S}$cience). It consists of 32 questions and a paired set of guidelines. REFORMS was developed based on a consensus of 19 researchers across computer science, data science, mathematics, social sciences, and biomedical sciences. REFORMS can serve as a resource for researchers when designing and implementing a study, for referees when reviewing papers, and for journals when enforcing standards for transparency and reproducibility.

cs.LG↗

arXiv2022

Security policy audits: why and how

Information security isn't just about software and hardware -- it's at least as much about policies and processes. But the research community overwhelmingly focuses on the former over the latter, while gaping policy and process problems persist. In this experience paper, we describe a series of security policy audits that we conducted, exposing policy flaws affecting billions of users that can be -- and often are -- exploited by low-tech attackers who don't need to use any tools or exploit software vulnerabilities. The solutions, in turn, need to be policy-based. We advocate for the study of policies and processes, point out its intellectual and practical challenges, lay out our theory of change, and present a research agenda.

cs.CR↗

arXiv2022

Leakage and the Reproducibility Crisis in ML-based Science

The use of machine learning (ML) methods for prediction and forecasting has become widespread across the quantitative sciences. However, there are many known methodological pitfalls, including data leakage, in ML-based science. In this paper, we systematically investigate reproducibility issues in ML-based science. We show that data leakage is indeed a widespread problem and has led to severe reproducibility failures. Specifically, through a survey of literature in research communities that adopted ML methods, we find 17 fields where errors have been found, collectively affecting 329 papers and in some cases leading to wildly overoptimistic conclusions. Based on our survey, we present a fine-grained taxonomy of 8 types of leakage that range from textbook errors to open research problems. We argue for fundamental methodological changes to ML-based science so that cases of leakage can be caught before publication. To that end, we propose model info sheets for reporting scientific claims based on ML models that would address all types of leakage identified in our survey. To investigate the impact of reproducibility errors and the efficacy of model info sheets, we undertake a reproducibility study in a field where complex ML models are believed to vastly outperform older statistical models such as Logistic Regression (LR): civil war prediction. We find that all papers claiming the superior performance of complex ML models compared to LR models fail to reproduce due to data leakage, and complex ML models don't perform substantively better than decades-old LR models. While none of these errors could have been caught by reading the papers, model info sheets would enable the detection of leakage in each case.

cs.LG↗

arXiv2022

How Algorithms Shape the Distribution of Political Advertising: Case Studies of Facebook, Google, and TikTok

Online platforms play an increasingly important role in shaping democracy by influencing the distribution of political information to the electorate. In recent years, political campaigns have spent heavily on the platforms' algorithmic tools to target voters with online advertising. While the public interest in understanding how platforms perform the task of shaping the political discourse has never been higher, the efforts of the major platforms to make the necessary disclosures to understand their practices falls woefully short. In this study, we collect and analyze a dataset containing over 800,000 ads and 2.5 million videos about the 2020 U.S. presidential election from Facebook, Google, and TikTok. We conduct the first large scale data analysis of public data to critically evaluate how these platforms amplified or moderated the distribution of political advertisements. We conclude with recommendations for how to improve the disclosures so that the public can hold the platforms and political advertisers accountable.

cs.SI↗

arXiv2022

Resurrecting Address Clustering in Bitcoin

Blockchain analysis is essential for understanding how cryptocurrencies like Bitcoin are used in practice, and address clustering is a cornerstone of blockchain analysis. However, current techniques rely on heuristics that have not been rigorously evaluated or optimized. In this paper, we tackle several challenges of change address identification and clustering. First, we build a ground truth set of transactions with known change from the Bitcoin blockchain that can be used to validate the efficacy of individual change address detection heuristics. Equipped with this data set, we develop new techniques to predict change outputs with low false positive rates. After applying our prediction model to the Bitcoin blockchain, we analyze the resulting clustering and develop ways to detect and prevent cluster collapse. Finally, we assess the impact our enhanced clustering has on two exemplary applications.

cs.CR↗

arXiv2022

The worst of both worlds: A comparative analysis of errors in learning from data in psychology and machine learning

Recent arguments that machine learning (ML) is facing a reproducibility and replication crisis suggest that some published claims in ML research cannot be taken at face value. These concerns inspire analogies to the replication crisis affecting the social and medical sciences. They also inspire calls for the integration of statistical approaches to causal inference and predictive modeling. A deeper understanding of what reproducibility concerns in supervised ML research have in common with the replication crisis in experimental science puts the new concerns in perspective, and helps researchers avoid "the worst of both worlds," where ML researchers begin borrowing methodologies from explanatory modeling without understanding their limitations and vice versa. We contribute a comparative analysis of concerns about inductive learning that arise in causal attribution as exemplified in psychology versus predictive modeling as exemplified in ML. We identify themes that re-occur in reform discussions, like overreliance on asymptotic theory and non-credible beliefs about real-world data generating processes. We argue that in both fields, claims from learning are implied to generalize outside the specific environment studied (e.g., the input dataset or subject sample, modeling implementation, etc.) but are often impossible to refute due to undisclosed sources of variance in the learning pipeline. In particular, errors being acknowledged in ML expose cracks in long-held beliefs that optimizing predictive accuracy using huge datasets absolves one from having to consider a true data generating process or formally represent uncertainty in performance claims. We conclude by discussing risks that arise when sources of errors are misdiagnosed and the need to acknowledge the role of human inductive biases in learning and reform.

cs.LG↗

arXiv2021

Mitigating Dataset Harms Requires Stewardship: Lessons from 1000 Papers

Machine learning datasets have elicited concerns about privacy, bias, and unethical applications, leading to the retraction of prominent datasets such as DukeMTMC, MS-Celeb-1M, and Tiny Images. In response, the machine learning community has called for higher ethical standards in dataset creation. To help inform these efforts, we studied three influential but ethically problematic face and person recognition datasets -- Labeled Faces in the Wild (LFW), MS-Celeb-1M, and DukeMTM -- by analyzing nearly 1000 papers that cite them. We found that the creation of derivative datasets and models, broader technological and social change, the lack of clarity of licenses, and dataset management practices can introduce a wide range of ethical concerns. We conclude by suggesting a distributed approach to harm mitigation that considers the entire life cycle of a dataset.

cs.LG↗

arXiv2021

The Impact of User Location on Cookie Notices (Inside and Outside of the European Union)

The web is global, but privacy laws differ by country. Which set of privacy rules do websites follow? We empirically study this question by detecting and analyzing cookie notices in an automated way. We crawl 1,500 European, American, and Canadian websites from each of 18 countries. We detect cookie notices on 40 percent of websites in our sample. We treat the presence or absence of cookie notices, as well as visual differences, as proxies for differences in privacy rules. Using a series of regression models, we find that the website's Top Level Domain explains a substantial portion of the variance in cookie notice metrics, but the user's vantage point does not. This suggests that websites follow one set of privacy rules for all their users. There is one exception to this finding: cookie notices differ when accessing .com domains from inside versus outside of the EU. We highlight ways in which future research could build on our preliminary findings.

cs.CY↗

arXiv2021

Simulation as Experiment: An Empirical Critique of Simulation Research on Recommender Systems

Simulation can enable the study of recommender system (RS) evolution while circumventing many of the issues of empirical longitudinal studies; simulations are comparatively easier to implement, are highly controlled, and pose no ethical risk to human participants. How simulation can best contribute to scientific insight about RS alongside qualitative and quantitative empirical approaches is an open question. Philosophers and researchers have long debated the epistemological nature of simulation compared to wholly theoretical or empirical methods. Simulation is often implicitly or explicitly conceptualized as occupying a middle ground between empirical and theoretical approaches, allowing researchers to realize the benefits of both. However, what is often ignored in such arguments is that without firm grounding in any single methodological tradition, simulation studies have no agreed upon scientific norms or standards, resulting in a patchwork of theoretical motivations, approaches, and implementations that are difficult to reconcile. In this position paper, we argue that simulation studies of RS are conceptually similar to empirical experimental approaches and therefore can be evaluated using the standards of empirical research methods. Using this empirical lens, we argue that the combination of high heterogeneity in approaches and low transparency in methods in simulation studies of RS has limited their interpretability, generalizability, and replicability. We contend that by adopting standards and practices common in empirical disciplines, simulation researchers can mitigate many of these weaknesses.

cs.CY↗

arXiv2021

T-RECS: A Simulation Tool to Study the Societal Impact of Recommender Systems

Simulation has emerged as a popular method to study the long-term societal consequences of recommender systems. This approach allows researchers to specify their theoretical model explicitly and observe the evolution of system-level outcomes over time. However, performing simulation-based studies often requires researchers to build their own simulation environments from the ground up, which creates a high barrier to entry, introduces room for implementation error, and makes it difficult to disentangle whether observed outcomes are due to the model or the implementation. We introduce T-RECS, an open-sourced Python package designed for researchers to simulate recommendation systems and other types of sociotechnical systems in which an algorithm mediates the interactions between multiple stakeholders, such as users and content creators. To demonstrate the flexibility of T-RECS, we perform a replication of two prior simulation-based research on sociotechnical systems. We additionally show how T-RECS can be used to generate novel insights with minimal overhead. Our tool promotes reproducibility in this area of research, provides a unified language for simulating sociotechnical systems, and removes the friction of implementing simulations from scratch.

cs.CY↗