Pasar al contenido principal

ROBUST INTELLIGENT REASONING: TECHNIQUES, EVALUATION AND APPLICATIONS

PID2024-162030OB-I00

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para la Investigación y el Desarrollo Experimental
Subprograma Subprograma Estatal de Generación de Conocimiento Científico-Técnico y Desarrollo Experimental
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2024
Plan Estatal 2024-2027 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 7
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

General scales unlock AI evaluation with explanatory and predictive power

  • Zhou, Lexin
  • Pacchiardi, Lorenzo
  • Collins, Katherine M.
  • Zhang, Seraphina
  • Zhao, Qinlin
  • Huang, Yitian
  • Sun, Luning
  • Prunty, Jonathan E.
  • Li, Zongqian
  • Sanchez-Garcia, Pablo
  • Casares, Pablo A. M.
  • Zu, Jiyun
  • Martínez-Plumed, Fernando|||0000-0003-2902-6477
  • Moros-Daval, Yael
  • Jiang-Chen, Ke-Xin
  • Burden, John
  • Mehrbakhsh, Behzad|||0000-0001-9017-989X
  • Stillwell, David
  • Cebrian, Manuel
  • Wang, Jindong
  • Henderson, Peter
  • Wu, Sherry Tongshuang
  • Kyllonen, Patrick C.
  • Cheke, Lucy
  • Xie, Xing
  • Hernández-Orallo, José|||0000-0001-9746-7632
[EN] Ensuring safe and effective use of artificial intelligence (AI) requires understanding and anticipating its performance on new tasks, from advanced scientific challenges to transformed workplace activities1, 2-3. So far, benchmarking has guided progress in AI but has offered limited explanatory and predictive power for general-purpose AI systems4, 5, 6, 7-8, attributed to limited transferability across specific tasks9, 10-11. Here we introduce general scales for AI evaluation that elicit demand profiles explaining what capabilities common AI benchmarks truly measure, extract ability profiles quantifying the general strengths and limits of AI systems and robustly predict AI performance for new task instances. Our fully automated methodology builds on 18 rubrics, capturing a broad range of cognitive and intellectual demands, which place different task instances on the same general scales, illustrated on 15 large language models (LLMs) and 63 tasks. Both the demand and the ability profiles on these scales bring new insights such as construct validity through benchmark sensitivity and specificity and explain conflicting claims about whether AI has reasoning capabilities. Ultimately, high predictive power at the instance level becomes possible using the general scales, providing superior estimates over strong black-box baseline predictors, especially in out-of-distribution settings (new tasks and benchmarks). The scales, rubrics, battery, techniques and results presented here constitute a solid foundation for a science of AI evaluation, underpinning the reliable deployment of AI in the years ahead., We thank the POLARIS Lab and the Computational Cognitive Science Lab at Princeton University, the European AI Office, the UK AISI, the Future of Life Institute, the OECD Future of Skills team, S. Chang, M. Zilka, J. Lian, R. Burnell, W. Schellaert, A. Gomez, M. Tesic, C. Li and P. Romero for their valuable help and feedback at certain stages of the project. We thank OpenAI for granting us research access to several LLMs to conduct the experiments in this paper and DeepSeek and Meta for giving us access to the weights of their models. We acknowledge support from the following institutions: Microsoft Accelerate Foundation Models Research (AFMR) grant programme, Long-Term Future Scholarship financed by Coefficient Giving (formerly Open Philanthropy) and the Spanish Government's Knowledge Generation Projects (PID2023-150271NB-C21). This work was also supported by CIPROM/2022/6 (FASSLOW), IDIFEDER/2021/05 (CLUSTERIA), CIACIF/2023/276, financed by Generalitat Valenciana, the EC H2020-EU grant agreement no. 952215 (TAILOR), Spanish grants PID2021-122830OB-C42 (SFERA), PID2023-150271NB-C21 and PID2024-162030OB-100 (ROBIN) financed by MCIN/AEI/10.13039/501100011033 and 'ERDF A way of making Europe', Catedra ENIA-UPV in Sustainable AI Development, TSI-100930-2023-9, INCIBE's Chair financed by the EU's NextGenerationEU, EUR2024-153548 (PREDAIT) 'Towards Predictable AI' from 'Spanish Europe Excelencia' 2024, Spanish National Research Council (CSIC) Special Intramural Projects programme and the Cambridge Trust. M.C. declares support from Google.org through the Silicon Valley Community Foundation by means of a grant to Fundacion General CSIC. The research of J.H.-O. is supported by OpenAI's grant to the 'AI Progress through the Lens of Predictable AI Ecosystems' programme, which is based at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Robustness under noise: assessing the impact of perturbed key attributes on machine learning models

  • Padró-Ferragut, Cristina
  • Martínez-Plumed, Fernando|||0000-0003-2902-6477
  • Ramírez Quintana, María José|||0000-0002-0559-3568
[EN] Machine learning models often struggle when confronted with the complexities of real-world data, such as noise and incomplete information, which pose significant challenges for practical applications. This paper presents a systematic framework for assessing the robustness of machine learning models against these adversities. In particular, we investigate how noise affecting the most critical attribute of a dataset impacts on model predictions. Our methodology uses a variety of datasets, a wide range of machine learning models, and a customizable noise induction function, allowing for comprehensive evaluations tailored to specific application needs. Through an extensive experimental analysis, we identify the vulnerabilities of models when their critical attributes are perturbed. In addition, by applying hierarchical clustering, we categorise machine learning families based on their robustness, establishing a taxonomy that highlights the susceptibility of different models to noise. Our findings provide important insights into the interplay between model performance and noise levels and improve the understanding of model behaviour under adverse conditions., We acknowledge support from: CIPROM/2022/6 (FASSLOW) funded by Generalitat Valenciana; and Spanish grant PID2024-162030OB-I00 (ROBIN) funded by MCIN/AEI/10.13039/ 501100011033 and ERDF A way of making Europe", Cátedra ENIA-UPV in Sustainable AI Development, TSI-100930-2023-9, and INCIBE s Chair funded by the EU-NextGenerationEU through the Spanish government s Plan de Recuperación, Transformación y Resiliencia. CPF is supported by UPV under FPI grant PAID-01-22. Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. Funding for open access charge: CRUE-Universitat Politecnica de Valencia.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Predictable artificial intelligence

  • Zhou, Lexin
  • Casares, Pablo A. M.
  • Burden, John
  • Burnell, Ryan
  • Cheke, Lucy
  • Marcoci, Alexandru
  • Heigeartaigh, Sean O.
  • Rutar, Danaja
  • Schellaert, Wout Willy M.
  • Voudouris, Konstantinos
  • Martínez-Plumed, Fernando|||0000-0003-2902-6477
  • Ferri Ramírez, César|||0000-0002-8975-1120
  • Mehrbakhsh, Behzad|||0000-0001-9017-989X
  • Moros-Daval, Yael
  • Hernández-Orallo, José|||0000-0001-9746-7632
[EN] Many areas of artificial intelligence, and machine learning in particular, aim at being probably correct, i.e., valid on average, rather than pursuing the idealistic goal of being provably valid for all inputs. However, AI systems could still be predictably valid, such as an imperfect robot deliverer for which we can reliably and precisely predict the task instances for which it is correct and safe, its valid operating range. "Predictable AI" is a nascent research area that explores ways of anticipating key validity indicators (e.g., performance, safety) of present and future AI ecosystems. We argue that achieving predictability is crucial for fostering trust, liability, control, alignment and safety of AI, and thus should be prioritised over performance. We formally characterise predictability, explore its most relevant components, illustrate what can be predicted, describe alternative candidates for predictors, as well as the trade-offs between maximising validity and predictability. To illustrate these concepts, we bring an array of illustrative examples covering diverse ecosystem configurations. "Predictable AI" is related to other areas of technical and non-technical AI research, but have distinctive questions, hypotheses, techniques and challenges. This paper aims to elucidate them, calls for identifying paths towards a landscape of predictably valid AI systems and outlines the potential impact of this emergent field., This work has been supported by Open Philanthropy Long-term Future Fund, CIPROM/2022/6 (FASSLOW) , funded by Generalitat Valenciana, and Spanish grant PID2024-162030OB-100 (ROBIN) , funded by MCIN/AEI/10.13039/501100011033 and ERDF A way of making Europe, Catedra ENIA-UPV in Sustainable AI Development, TSI-100930-2023-9, and INCIBE's Chair funded by the EU-NextGenerationEU through the Spanish government's Plan de Recuperacion, Transformacion y Resiliencia, and EUR2024-153548 (PREDAIT) "Towards Predictable AI" from "Spanish Europe Excelencia" 2024.JHO's research is supported by OpenAI's grant to the 'AI Progress through the Lens of Predictable AI Ecosystems' programme, which is based at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

DM-Check: Verifying invariants of concurrent systems by deductive model checking

  • Bae, Kyungmin
  • Meseguer, Jose
  • Escobar Román, Santiago|||0000-0002-3550-4781
  • López-Rueda, Raúl|||0000-0002-2550-8041
  • Sapiña-Sanchis, Julia|||0000-0003-2994-6986
[EN] We propose a new deductive model checking methodology where narrowing-based logical model checking of symbolic states specified as disjunctions of constrained patterns is combined with inductive theorem proving to discharge inductive verification conditions that ensure useful symbolic state space reductions. An obvious combination is to use an inductive theorem prover in automated mode as an oracle to help logical model checking reach a fixpoint. But this is not the only possible combination. In this paper we focus instead on a new deductive model checking methodology to verify invariants -including inductive invariants- of infinite-state systems, where logical model checking automates large parts of the verification effort with the help of an inductive theorem prover as an oracle. Inductive verification conditions not discharged automatically by the oracle are dealt with by commands that refine some constrained patterns by useful semantic equivalences, and by using an inductive theorem prover in interactive mode. This methodology is demonstrated by means of concurrent system examples using two Maude tools working in tandem: the DM-Check narrowing-based symbolic model checker, and the NuITP inductive theorem prover., S. Escobar, J. Sapina, and R. Lopez have been partially supported by the grant CIPROM/2022/6 funded by Generalitat Valenciana, by the grant PID2024-162030OB-100 funded by MCIN/AEI/10.13039/501100011033 and ERDF A way of making Europe, and byINCIBE's Chair funded by the EU-NextGenerationEU through the Spanish government's Plan de Recuperacion, Transformacion y Resiliencia. K. Bae, S. Escobar, J. Sapina, and R. Lopez have been supported by the NATO Science for Peace and Security Programme project SymSafe (grant number G6133) . K. Bae was supported by the National Research Foundation of Korea (NRF) grant (No. RS-2021-NR060080) and the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant (No. RS-2024-00439856) , both funded by the Korea government (MSIT) .r INCIBE's Chair funded by the EU-NextGenerationEU through the Spanish government's Plan de Recuperacion, Transformacion y Resiliencia. K. Bae, S. Escobar, J. Sapina, and R. Lopez have been supported by the NATO Science for Peace and Security Programme project SymSafe (grant number G6133) . K. Bae was supported by the National Research Foundation of Korea (NRF) grant (No. RS-2021-NR060080) and the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant (No. RS-2024-00439856) , both funded by the Korea government (MSIT) .
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

NuITP: Accelerating the inductive verification of equational programs through symbolic simplification

  • Duran, Francisco
  • Meseguer, Jose
  • Escobar Román, Santiago|||0000-0002-3550-4781
  • Sapiña-Sanchis, Julia|||0000-0003-2994-6986
[EN] NuITP is an inductive theorem prover that combines advanced symbolic techniques such as narrowing, equality predicates, variant unification, variant satisfiability, order-sorted congruence closure, ordered rewriting, and strategy-based rewriting (all applied modulo axioms) to verify equational programs with expressive features such as sorts and subsorts, conditional equations and rewriting modulo axioms in Maude and in other equational languages. This paper presents the tool, explains its inference rules, and illustrates their use in proving several examples with substantial levels of automated proof simplification based on symbolic techniques., We cordially thank Claude Kirchner for his comments, which have helped us improve the paper. F. Duran has been partially supported by Spanish Government project PID2025-172454NB-I00. S. Escobar and J. Sapina have been partially supported by the grant CIPROM/2022/6 funded by Generalitat Valenciana and by the grant PID2024-162030OB-100 funded by MCIN/AEI/10.13039/501100011033 and ERDF A way of making Europe.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Should computations halt?

  • Lucas Alba, Salvador|||0000-0001-9923-2108
[EN] The so-called Hilbert's Program, aiming at a full axiomatization of mathematics, fuelled the quest to provide an answer to the Entscheidungsproblem, the Decision Problem, aiming to find effective methods to obtain the truth value of logical sentences. Intuitively, when using such methods a halting behavior would be desirable, thus providing a clear answer to the prospective user. This was implicit in Hilbert's finitist approach to the foundations of mathematics and practical use of logical methods. In 1936 the quest for such effective methods brought four different approaches: Kleene's formulation of G & ouml;del/Herbrand general recursion, Church's lambda-definability, Turing's computation with automatic machines (a-machines), and Post's machines. However, the halting issue was treated quite differently in those approaches. In particular, the technical meaning of "computation" was explicitly introduced by Turing in his 1936 paper and he payed no attention to halting computations there. This paper traces how the halting requirement became a stable part of the definition of computation which is often used today., Supported by the EU (FEDER), project PID2024-162030OB-100 funded by MCIN/AEI/10.13039/501100011033 and by "ERDF A way of making Europe" and by the grant CIPROM/2022/6 funded by Generalitat Valenciana.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Evaluating General-Purpose AI with Psychometrics

  • Wang, Xiting
  • Jiang, Liming
  • Stillwell, David
  • Chen, Shiqiang
  • Sun, Luning
  • Luo, Fang
  • Xie, Xing
  • Hernández-Orallo, José|||0000-0001-9746-7632
[EN] IN THE RAPIDLY evolving field of artificial intelligence (AI), rigorous evaluation is critical. Evaluation helps prevent disastrous outcomes in high-stakes applications such as autonomous driving and medical diagnosis,7 and helps stop the spread of racist, sexist, ableist, extremist, and other harmful ideologies.4 It also plays a key role in understanding how intelligent AI is, preventing the misallocation of resources, and guiding improvements in future model development. Despite its importance, rigorous evaluation of versatile general-purpose AI systems such as ChatGPT and Gemini is challenging. The versatility of these systems, which can handle diverse tasks specified, This work was supervised by Luning Sun, Fang Luo, and Xing Xie. The authors would like to thank Jinyan Fan, Marija Slavkovik, Clemens Stachl, Alina A von Davier, Xiangen Hu, Yu Lu, Bryan Maddox, Mengxiao Zhu, Meng Li, Li Dong, Jindong Wang, Igor Sterner, Greg Serapio-Garcia, Peter Romero, Fengli Xu, Fernando Martinez Plumed, and Lidong Zhou for the discussions and comments on early versions of the manuscript. This research was funded by Beijing Natural Science Foundation L247011; Microsoft Research Asia Collaborative Research Program grant FY23-Research-Sponsorship-422 "The convergence of assessing human and big model capabilities"; CIPROM/2022/6 (FASSLOW) funded by Generalitat Valenciana, and Spanish grant PID2024-162030OB-100 (ROBIN) funded by MCIN/AEI/10.13039/501100011033 and ERDF A way of making Europe, Catedra ENIA-UPV in Sustainable AI Development, TSI-100930-2023-9, and INCIBE's Chair funded by the EU-NextGenerationEU through the Spanish government's Plan de Recuperacion, Transformacion y Resiliencia, and EUR2024-153548 (PREDAIT) "Towards Predictable AI" from "Spanish Europe Excelencia" 2024; National Natural Science Foundation of China (NSFC) (NO. 62476279, NO. 92470205, NO. U2436209); Major Innovation & Planning Interdisciplinary Platform for the "Double-First Class" Initiative, Renmin University of China; The Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China No. 24XNKJ18; Fund for building world-class universities (disciplines) of Renmin University of China and Public Computing Cloud, Renmin University of China. L.S. and D.S. gratefully acknowledge financial support from Invesco through their philanthropic donation to Cambridge Judge Business School.