90% of shrimp farms in South East Asia already have at least 1 dormant disease living in the water, waiting for a shrimp’s immune system to be compromised

Jala

This paper is available on arxiv under CC 4.0 license. Authors: (1) Ajay Krishnan T. K., School of Digital Sciences; (2) V. S. Anoop, School of Digital Sciences. Table of Links Abstract & Introduction Related Studies Materials and Methods Proposed Approach Results and Discussions Conclusions and References 6 Conclusions Climate change, a pressing global concern, necessitates thorough analysis and understanding across diverse domains to mitigate its impacts effectively. In recent years, the fusion of NLP techniques and machine learning algorithms has emerged as a promising approach for comprehending the complexities and nuances of climate change through the lens of textual data. This paper utilized the advancements in domain-specific large language models to harness the potential of NLP in addressing the challenges posed by climate change through sentiment analysis. By leveraging advanced NLP methodologies, we could identify climate change discourse that may enable uncovering valuable insights and facilitate informed decision-making. References VS Anoop. Sentiment classification of diabetes-related tweets using transformer-based deep learning approach. In International Conference on Advances in Computing and Data Sciences, pages 203–214. Springer, 2023. VS Anoop and S Sreelakshmi. Public discourse and sentiment during mpox outbreak: an analysis using natural language processing. Public Health, 218:114–120, 2023. VS Anoop, Jose Thekkiniath, and Usharani Hareesh Govindarajan. We chased covid-19; did we forget measles?-public discourse and sentiment analysis on spiking measles cases using natural language processing. In International Conference on Multi-disciplinary Trends in Artificial Intelligence, pages 147–158. Springer, 2023. Sina Ardabili, Amir Mosavi, Majid Dehghani, and Annamária R Várkonyi-Kóczy. Deep learning and machine learning in hydrological processes climate change and earth systems a systematic review. In Engineering for Sustainable Future: Selected papers of the 18th International Conference on Global Research and Education Inter-Academia– 2019 18, pages 52–62. Springer, 2020. Ceylan Ceylan. Application of Natural Language Processing to Unstructured Data: A Case Study of Climate Change. PhD thesis, Massachusetts Institute of Technology, 2022. B Jalalzadeh Fard, SA Hasan, and JE Bell. Climedbert: A pre-trained language model for climate and health-related text. arXiv preprint arXiv:2212.00689, 2022. Muhammad Khurram Iqbal, Kamran Abid, Salah u din Ayubi, Naeem Aslam, et al. Omicron tweet sentiment analysis using ensemble learning. Journal of Computing & Biomedical Informatics, 4(02):160–171, 2023. Sarin Jickson, VS Anoop, and S Asharaf. Machine learning approaches for detecting signs of depression from social media. In Proceedings of International Conference on Information Technology and Applications: ICITA 2022, pages 201–214. Springer, 2023. Romieo John, VS Anoop, and S Asharaf. Health mention classification from user-generated reviews using machine learning techniques. In Proceedings of International Conference on Information Technology and Applications: ICITA 2022, pages 175–188. Springer, 2023. S Lekshmi and VS Anoop. Sentiment analysis on covid-19 news videos using machine learning techniques. In Proceedings of International Conference on Frontiers in Computing and Systems: COMSYS 2021, pages 551–560. Springer, 2022. Mustapha Lydiri, Yousef El Mourabit, Youssef El Habouz, and Mohamed Fakir. A performant deep learning model for sentiment analysis of climate change. Social Network Analysis and Mining, 13(1):8, 2022. Nabila Mohamad Sham and Azlinah Mohamed. Climate change sentiment analysis using lexicon, machine learning and hybrid approaches. Sustainability, 14(8):4723, 2022. Uddagiri Sirisha and Sai Chandana Bolem. Aspect based sentiment & emotion analysis with roberta, lstm. International Journal of Advanced Computer Science and Applications, 13(11), 2022. Apoorva Upadhyaya, Marco Fisichella, and Wolfgang Nejdl. A multi-task model for sentiment aided stance detection of climate change tweets, 2022. Milan Varghese and VS Anoop. Deep learning-based sentiment analysis on covid-19 news videos. In Proceedings of International Conference on Information Technology and Applications: ICITA 2021, pages 229–238. Springer, 2022. Francesco S Varini, Jordan Boyd-Graber, Massimiliano Ciaramita, and Markus Leippold. Climatext: A dataset for climate change topic detection. arXiv preprint arXiv:2012.00483, 2020. Nicolas Webersinke, Mathias Kraus, Julia Anna Bingler, and Markus Leippold. Climatebert: A pretrained language model for climate-related text. arXiv preprint arXiv:2110.12010, 2021.

ClimateNLP: Analyzing Public Sentiment Towards Climate Change: Conclusions and References

This paper is under CC 4.0 license. available on arxiv Authors: (1) Marcin W ˛atorek, Faculty of Computer Science and Telecommunications, Cracow University of Technology, ul. Warszawska 24, 31-155 Kraków, Poland and marcin.watorek@pk.edu.pl (M.W.); (2) Jarosław Kwapie ´n, Complex Systems Theory Department, Institute of Nuclear Physics, Polish Academy of Sciences, Radzikowskiego 152, 31-342 Kraków, Poland and jaroslaw.kwapien@ifj.edu.pl; (3)  Stanisław Drozd˙ z, Faculty of Computer Science and Telecommunications, Cracow University of Technology, ul. Warszawska 24, 31-155 Kraków, Poland, Complex Systems Theory Department, Institute of Nuclear Physics, Polish Academy of Sciences, Radzikowskiego 152, 31-342 Kraków, Poland and stanislaw.drozdz@ifj.edu.pl (S.D.). Table of Links Abstract and Introduction Data and Methodology Results and Discussion Conclusion and References 4. Conclusions Based on the multiscale cross-correlation analysis performed for the data covering almost the last three years, it can be concluded that the cryptocurrency market dynamics is substantially tied to the traditional financial markets. Consistently, the most liquid cryptocurrencies, BTC and ETH, cannot serve as a hedge or safe haven for the stock market investments, not only during the turbulent periods like the Covid-19 panic, where this effect is particularly strong, but also during the recent bear market period on tech stocks, which has been accompanied by the parallel bear market on cryptocurrencies. Many observations show that the Covid-19 pandemic may have changed the paradigm that the cryptocurrency market is a largely autonomous market. The recent market developments and the strong US dollar have additionally increased the strength of the cross-correlations between BTC and ETH on the one side and the US tech stocks on the other side. These observations support some earlier findings on the same subject (e.g., [22,70]). In contrast, as the cryptocurrency market was weakly correlated with other markets during 2021, our results cannot support directly a recent hypothesis that the quantitative easing could actually be responsible for these correlations [22]. The existence of links between the global economy and the cryptocurrency market are further strengthened by the reaction of the price changes of BTC and ETH to economic data, such as CPI inflation, in a similar way to traditional financial instruments. These results are able to remove or, at least, to suppress the uncertainty that recent literature on this topic has brought to the cryptocurrency investors. Now it is more clear that the cryptocurrencies can no longer serve as a convenient hedging target for the investors whose purpose is to diversify the risk. Our study brings a strong indication that the cryptocurrency market has finally become a connected part of the global financial markets after 12 years of the maturation process. Whether such a direction of this market evolution remains in agreement with the early vision of the cryptocurrency creators can be disputed, however. We also face a related question: does the fact that we have got “just another part of the global financial market” deserve devoting so huge amounts of energy to it? Sooner or later this question must be addressed by the policy makers. Nevertheless, what becomes evident now is that it allows the market participants to broaden the spectrum of their investment possibilities. Among the limitations that might have influenced our study and, subsequently, our conclusions, we have to mention that only two principal cryptocurrencies were studied. Although they are the most influential, the most frequently traded, and widely discussed cryptocurrencies, they by no means define the entire market. It is possible that an analysis that included some less important cryptocurrencies would bring different outcomes. This is especially likely for the marginal cryptocurrencies without any thinkable “fundamental” value, whose dynamics is driven predominantly by extreme speculation. However, as the cryptocurrency market is looked at by the most investors through the lens of BTC and ETH (as their capitalization indicates), this particular limitation does not seem discouraging to us. Currently, these two assets shape the whole cryptocurrency market and we expect them to continue doing it in the nearest future. Another limitation of our analysis is the particular selection of the traditional financial instruments. Indeed, they constitute only a small fraction of the available ones. We are convinced, though, that they are among the most observed and the most influential ones in the context of the global economy, which fully justifies our choice. A more general observation that the cryptocurrency market has spontaneously coupled to the technological sector of the stock markets by reacting to some trigger provided by the external data inflow resembles analogous effects of the spontaneous emergence of order among the so-far independent degrees of freedom in the various complex systems. However, as complexity allows for flexible reacting of a system to both the external perturbations and internal processes, such effects of ordering in the financial markets have to be eventually counterbalanced by the opposite processes of disordering. Therefore, the market participants must be aware that the inter-market couplings may not last forever and they can significantly be weakened or even removed completely at some point in future. This is why the in-depth studies of the cross-market dependencies have to remain among the principal directions of the cryptocurrency research. Our future work will also deal with energy consumption of the cryptocurrency market. Conceptualisation, S.D., J.K. and M.W.; methodology, S.D., J.K. and M.W.; software, M.W.; validation, S.D., J.K. and M.W.; formal analysis, M.W.; investigation, S.D., J.K. and M.W.; resources, M.W.; data curation, M.W.; writing—original draft preparation, M.W.; writing review and editing, S.D., J.K. and M.W.; visualisation, M.W.; supervision, S.D. All authors have read and agreed to the published version of the manuscript. Author Contributions: This research received no external funding. Funding: Data available freely from Dukascopy [58]. Data Availability Statement: The authors declare no conflict of interest. Conflicts of Interest: References 1. Kwapie ´n, J.; Drozd˙ z, S. Physical approach to complex systems. ˙ Phys. Rep. 2012, 515, 115–226. 2. Bitcoin Pizzaday. Available online: https://www.investopedia.com/news/bitcoin-pizza-day-celebrating-20-million-pizza-order/ (accessed on 1 November 2022). 3. CoinMarketCap. Available online: https://coinmarketcap.com (accessed on 1 November 2022). 4. Gerlach, J.C.; Demos, G.; Sornette, D. Dissection of Bitcoin’s multiscale bubble history from January 2012 to February 2018. R. Soc. Open Sci. 2019, 6, 180643. 5. Bellon, C.; Figuerola-Ferretti, I. Bubbles in Ethereum. Financ. Res. Lett. 2022, 46, 102387. 6. Zitis, P.I.; Contoyiannis, Y.; Potirakis, S.M. Critical dynamics related to a recent Bitcoin crash. Int. Rev. Financ. Anal. 2022, 84, 102368. 7. Aste, T. Cryptocurrency market structure: Connecting emotions and economics. Digit. Financ. 2019, 1, 5–21. 8. Maouchi, Y.; Charfeddine, L.; Montasser, G.E. Understanding digital bubbles amidst the COVID 19 pandemic: Evidence from DeFi and NFTs. Financ. Res. Lett. 2022, 47, 102584. 9. W ˛atorek, M.; Drozd˙ z, S.; Kwapie ´n, J.J.; Minati, L.; O´swi ˛ecimka, P.; Stanuszek, M. Multiscale characteristics of the emerging global ˙ cryptocurrency market. Phys. Rep. 2021, 901, 1–82. 10. Lahmiri, S.; Bekiros, S.; Bezzina, F. Complexity analysis and forecasting of variations in cryptocurrency trading volume with support vector regression tuned by Bayesian optimization under different kernels: An empirical comparison from a large dataset. Expert Syst. Appl. 2022, 209, 118349. 11. Drozd˙ z, S.; G ˛ebarowski, R.; Minati, L.; O´swi ˛ecimka, P.; W ˛atorek, M. Bitcoin market route to maturity? Evidence from return ˙ fluctuations, temporal correlations and multiscaling effects. Chaos 2018, 28, 071101. 12. Drozd˙ z, S.; Minati, L.; O´swi ˛ecimka, P.; Stanuszek, M.; W ˛atorek, M. Competition of noise and collectivity in global cryptocurrency ˙ trading: Route to a self-contained market. Chaos 2020, 30, 023122. 13. Kaiser, L.; Stöckl, S. Cryptocurrencies: Herding and the transfer currency. Financ. Res. Lett. 2020, 33, 101214. 14. Aslanidis, N.; Bariviera, A.F.; Perez-Laborda, A. Are cryptocurrencies becoming more interconnected? Econ. Lett. 2021, 199, 109725. 15. Kwapie ´n, J.; W ˛atorek, M.; Drozd˙ z, S. Cryptocurrency Market Consolidation in 2020-2021. Entropy 2021, 23, 1674. 16. Bae, G.; Kim, J.H. Observing Cryptocurrencies through Robust Anomaly Scores. Entropy 2022, 24, 1643. 17. James, N.; Menzies, M. Collective correlations, dynamics, and behavioural inconsistencies of the cryptocurrency market over time. Nonlinear Dyn. 2022, 107, 4001–4017. 18. Kakinaka, S.; Umeno, K. Asymmetric volatility dynamics in cryptocurrency markets on multi time scales. Res. Int. Bus. Financ. 2022, 62, 101754. 19. Arouxet, M.B.; Bariviera, A.F.; Pastor, V.E.; Vampa, V. Covid-19 impact on cryptocurrencies: Evidence from a wavelet-based Hurst exponent. Phys. A 2022, 596, 127170. 20. Corbet, S.; Hou, Y.G.; Hu, Y.; Larkin, C.; Lucey, B.; Oxley, L. Cryptocurrency liquidity and volatility interrelationships during the COVID-19 pandemic. Financ. Res. Lett. 2022, 45, 102137. 21. Nguyen, A.P.N.; Mai, T.T.; Bezbradica, M.; Crane, M. The Cryptocurrency Market in Transition before and after COVID-19: An Opportunity for Investors? Entropy 2022, 24, 1317. 22. Kumar, A.; Iqbal, N.; Mitra, S.K.; Kristoufek, L.; Bouri, E. Connectedness among major cryptocurrencies in standard times and during the COVID-19 outbreak. J. Int. Financ. Mark. Institutions Money 2022, 77, 101523. 23. W ˛atorek, M.; Kwapie ´n, J.; Drozd˙ z, S. Multifractal cross-Correlations of bitcoin and ether trading Characteristics in the Post-COVID-19 Time. Future Internet 2022, 14, 215. 24. Corbet, S.; Meegan, A.; Larkin, C.; Lucey, B.; Yarovaya, L. Exploring the dynamic relationships between cryptocurrencies and other financial assets. Econ. Lett. 2018, 165, 28–34. 25. Ji, Q.; Bouri, E.; Gupta, R.; Roubaud, D. Network causality structures among Bitcoin and other financial assets: A directed acyclic graph approach. Q. Rev. Econ. Financ. 2018, 70, 203–213. 26. Drozd˙ z, S.; Minati, L.; O´swi ˛ecimka, P.; Stanuszek, M.; W ˛atorek, M. Signatures of the Crypto Currency Market Decoupling from the Forex. Future Internet 2019, 11, 154. 27. Gil-Alana, L.A.; Abakah, E.J.A.; Rojo, M.F.R. Cryptocurrencies and stock market indices. Are they related? Res. Int. Bus. Financ. 2020, 51, 101063. 28. Manavi, S.A.; Jafari, G.; Rouhani, S.; Ausloos, M. Demythifying the belief in cryptocurrencies decentralized aspects. A study of cryptocurrencies time cross-correlations with common currencies, commodities and financial indices. Phys. A 2020, 556, 124759. 29. Urquhart, A.; Zhang, H. Is Bitcoin a hedge or safe haven for currencies? An intraday analysis. Int. Rev. Financ. Anal. 2019, 63, 49–57. 30. Wang, P.; Zhang, W.; Li, X.; Shen, D. Is cryptocurrency a hedge or a safe haven for international indices? A comprehensive and dynamic perspective. Financ. Res. Lett. 2019, 31, 1–18. 31. Shahzad, S.J.H.; Bouri, E.; Roubaud, D.; Kristoufek, L.; Lucey, B. Is Bitcoin a better safe-haven investment than gold and commodities? Int. Rev. Financ. Anal. 2019, 63, 322–330. 32. Shahzad, S.J.H.; Bouri, E.; Roubaud, D.; Kristoufek, L. Safe haven, hedge and diversification for G7 stock markets: Gold versus bitcoin. Econ. Model. 2020, 87, 212–224. 33. Bouri, E.; Shahzad, S.J.H.; Roubaud, D.; Kristoufek, L.; Lucey, B. Bitcoin, gold, and commodities as safe havens for stocks:New insight through wavelet analysis. Q. Rev. Econ. Financ. 2020, 77, 156–164. 34. Thampanya, N.; Nasir, M.A.; Huynh, T.L.D. Asymmetric correlation and hedging effectiveness of gold & cryptocurrencies: From pre-industrial to the 4th industrial revolution. Technol. Forecast. Soc. Chang. 2020, 159, 120195. 35. Zhu, X.; Niu, Z.; Zhang, H.; Huang, J.; Zuo, X. Can gold and bitcoin hedge against the COVID-19 related news sentiment risk? New evidence from a NARDL approach. Resour. Policy 2022, 79, 103098. 36. Conlon, T.; McGee, R. Safe haven or risky hazard? Bitcoin during the Covid-19 bear market. Financ. Res. Lett. 2020, 35, 101607. 37. Kristoufek, L. Grandpa, Grandpa, Tell Me the One About Bitcoin Being a Safe Haven: New Evidence From the COVID-19 Pandemic. Front. Phys. 2020, 8, 296. 38. Grobys, K. When Bitcoin has the flu: On Bitcoin’s performance to hedge equity risk in the early wake of the COVID-19 outbreak. Appl. Econ. Lett. 2021, 28, 860–865. 39. James, N.; Menzies, M.; Chan, J. Changes to the extreme and erratic behaviour of cryptocurrencies during COVID-19. Phys. A 2021, 565, 125581. 40. Barbu, T.C.; Boitan, I.A.; Cepoi, C.. Are cryptocurrencies safe havens during the COVID-19 pandemic? A threshold regression perspective with pandemic-related benchmarks. Econ. Bus. Rev. 2022, 8, 22–49. 41. Jareno, F.; Gonzàlez, M.D.L.O.; Belmonte, P. Asymmetric interdependencies between cryptocurrency and commodity markets: The COVID-19 pandemic impact. Quant. Financ. Econ. 2022, 6, 83–112. 42. Kakinaka, S.; Umeno, K. Cryptocurrency market efficiency in short- and long-term horizons during COVID-19: An asymmetric multifractal analysis approach. Financ. Res. Lett. 2022, 46, 102319. 43. Evrim Mandaci, P.; Cagli, E.C. Herding intensity and volatility in cryptocurrency markets during the COVID-19. Financ. Res. Lett. 2022, 46, 102382. 44. Baker, S.R.; Bloom, N.; Davis, S.J.; Kost, K.; Sammon, M.; Viratyosin, T. The Unprecedented Stock Market Reaction to COVID-19. Rev. Asset Pricing Stud. 2020, 10, 742–758. 45. Yarovaya, L.; Matkovskyy, R.; Jalan, A. The COVID-19 black swan crisis: Reaction and recovery of various financial markets. Res. Int. Bus. Financ. 2022, 59, 101521. 46. Conlon, T.; Corbet, S.; McGee, R.J. Are cryptocurrencies a safe haven for equity markets? An international perspective from the COVID-19 pandemic. Res. Int. Bus. Financ. 2020, 54, 101248. 47. Drozd˙ z, S.; Kwapie ´n, J.; O´swi ˛ecimka, P.; Stanisz, T.; W ˛atorek, M. Complexity in economic and social systems: Cryptocurrency market at around COVID-19. Entropy 2020, 22, 1043. 48. Caferra, R.; Vidal-Tomas, D. Who raised from the abyss? A comparison between cryptocurrency and stock market dynamics during the COVID-19 pandemic. Financ. Res. Lett. 2021, 43, 101954. 49. Jiang, Y.; Lie, J.; Wang, J.; Mu, J. Revisiting the roles of cryptocurrencies in stock markets: A quantile coherency perspective. Econ. Model. 2021, 95, 21–34. 50. James, N. Dynamics, behaviours, and anomaly persistence in cryptocurrencies and equities surrounding COVID-19. Phys. A 2021, 570, 125831. 51. Elmelki, A.; Chaâbane, N.; Benammar, R. Exploring the relationship between cryptocurrency and S&P500: evidence from wavelet coherence analysis. Int. J. Blockchains Cryptocurrencies 2022, 3, 256–268. 52. Wang, P.; Liu, X.; Wu, S. Dynamic Linkage between Bitcoin and Traditional Financial Assets: A Comparative Analysis of Different Time Frequencies. Entropy 2022, 24, 1565. 53. Balcilar, M.; Ozdemir, H.; Agan, B. Effects of COVID-19 on cryptocurrency and emerging market connectedness: Empirical evidence from quantile, frequency, and lasso networks. Phys. A 2022, 604, 127885. 54. Conlon, T.; Corbet, S.; McGee, R.J. Inflation and cryptocurrencies revisited: A time-scale analysis. Econ. Lett. 2021, 206, 109996. 55. Choi, S.; Shin, J. Bitcoin: An inflation hedge but not a safe haven. Financ. Res. Lett. 2022, 46, 102379. 56. James, N.; Menzies, M.; Chin, K. Economic state classification and portfolio optimisation with application to stagflationary environments. Chaos, Solitons Fractals 2022, 164, 112664. 57. Phochanachan, P.; Pirabun, N.; Leurcharusmee, S.; Yamaka, W. Do Bitcoin and Traditional Financial Assets Act as an Inflation Hedge during Stable and Turbulent Markets? Evidence from High Cryptocurrency Adoption Countries. Axioms 2022, 11, 339. 58. Dukascopy. Available online: https://www.dukascopy.com/swiss/pl/cfd/range-of-markets/ (accessed on 1 November 2022). 59. Pearson, K. Note on regression and inheritance in the case of two parents. Proc. R. Soc. Lond. 1895, 58, 240–242. 60. Gopikrishnan, P.; Plerou, V.; Nunes Amaral, L.A.; Meyer, M.; Stanley, H.E. Scaling of the distribution of fluctuations of financial market indices. Phys. Rev. E 1999, 60, 5305–5316. 61. Cont, R. Empirical properties of asset returns: Stylized facts and statistical issues. Quant. Financ. 2001, 1, 223–236. 62. Gabaix, X.; Gopikrishnan, P.; Plerou, V.; Stanley, H.E. A theory of power-law distributions in financial market fluctuations. Nature 2003, 423, 267–270. 63. Kwapie ´n, J.; O´swi ˛ecimka, P.; Drozd˙ z, S. Detrended fluctuation analysis made flexible to detect range of cross-correlated fluctuations. Phys. Rev. E 2015, 92, 052815. 64. Peng, C.K.; Buldyrev, S.V.; Havlin, S.; Simons, M.; Stanley, H.E.; Goldberger, A.L. Mosaic organization of DNA nucleotides. Phys. Rev. E 1994, 49, 1685–1689. 65. Jiang, Z.Q.; Xie, W.J.; Zhou, W.X.; Sornette, D. Multifractal analysis of financial markets: A review. Rep. Prog. Phys. 2019, 82, 125901. 66. O´swi ˛ecimka, P.; Drozd˙ z, S.; Kwapie ´n, J.; Górski, A. Effect of detrending on multifractal characteristics. ˙ Acta Phys. Pol. A 2013, 123, 597–603. 67. Zebende, G. DCCA cross-correlation coefficient: Quantifying level of cross-correlation. Phys. A 2011, 390, 614–618. 68. G ˛ebarowski, R.; O´swi ˛ecimka, P.; W ˛atorek, M.; Drozd˙ z, S. Detecting correlations and triangular arbitrage opportunities in the Forex by means of multifractal detrended cross-correlations analysis. Nonlinear Dyn. 2019, 98, 2349–2364. 69. W ˛atorek, M.; Drozd˙ z, S.; O´swi ˛ecimka, P.; Stanuszek, M. Multifractal cross-correlations between the world oil and other financial markets in 2012–2017. Energy Econ. 2019, 81, 874–885. 70. Zhang, Y.J.; Bouri, E.; Gupta, R.; Ma, S.J. Risk spillover between Bitcoin and conventional financial markets: An expectile-based approach. North Am. J. Econ. Financ. 2021, 55, 101296.

Integration of Cryptocurrency into the Global Financial Market: Conclusion and References

This paper is available on arxiv under CC 4.0 license. . Authors: Gemini Team, Google Table of Links Abstract and Introduction Model Architecture Training Infrastructure Training Dataset Evaluation Responsible Deployment Discussion and Conclusion, References Contributions and Acknowledgments Appendix 8. Contributions and Acknowledgments Leads Rohan Anil, Co-Lead, Text Sebastian Borgeaud, Co-Lead, Text Yonghui Wu, Co-Lead, Text Jean-Baptiste Alayrac, Co-Lead, MM Vision Jiahui Yu, Co-Lead, MM Vision Radu Soricut*, Co-Lead, MM Vision* Johan Schalkwyk, Lead, MM Audio Andrew M. Dai, Co-Lead, Data Anja Hauth*, Co-Lead, Data* Katie Millican, Co-Lead, Data David Silver, Co-Lead, Fine-Tuning Slav Petrov, Co-Lead, Fine-Tuning Melvin Johnson, Lead, Instruction Tuning Ioannis Antonoglou*, Co-Lead, RL Techniques* Julian Schrittwieser, Co-Lead, RL Techniques Amelia Glaese, Lead, Human Data Jilin Chen, Lead, Safety Emily Pitler, Co-Lead, Tool Use Timothy Lillicrap, Co-Lead, Tool Use Angeliki Lazaridou, Co-Lead, Eval Orhan Firat, Co-Lead, Eval James Molloy, Co-Lead, Infra Michael Isard, Co-Lead, Infra Paul R. Barham, Co-Lead, Infra Tom Hennigan, Co-Lead, Infra Benjamin Lee, Co-Lead, Codebase & Parallelism Fabio Viola, Co-Lead, Codebase & Parallelism Malcolm Reynolds, Co-Lead, Codebase & Parallelism Yuanzhong Xu, Co-Lead, Codebase & Parallelism Ryan Doherty, Lead, Ecosystem Eli Collins, Lead, Product Clemens Meyer, Co-Lead, Operations Eliza Rutherford, Co-Lead, Operations Erica Moreira, Co-Lead, Operations Kareem Ayoub, Co-Lead, Operations Megha Goel, Co-Lead, Operations Core Contributors George Tucker Enrique Piqueras Maxim Krikun Iain Barr Nikolay Savinov Ivo Danihelka Becca Roelofs Core Contributors Anaïs White Anders Andreassen Tamara von Glehn Lakshman Yagati Mehran Kazemi Lucas Gonzalez Misha Khalman Jakub Sygnowski Alexandre Frechette Charlotte Smith Laura Culp Lev Proleev Yi Luan Xi Chen James Lottes Nathan Schucher Federico Lebron Alban Rrustemi Natalie Clay Phil Crone Tomas Kocisky Jeffrey Zhao Bartek Perz Dian Yu Heidi Howard Adam Bloniarz Jack W. Rae Han Lu Laurent Sifre Marcello Maggioni Fred Alcober Dan Garrette Megan Barnes Shantanu Thakoor Jacob Austin Gabriel Barth-Maron William Wong Rishabh Joshi Rahma Chaabouni Deeni Fatiha Arun Ahuja Ruibo Liu Yunxuan Li Sarah Cogan Jeremy Chen Core Contributors Chao Jia Chenjie Gu Qiao Zhang Jordan Grimstad Ale Jakse Hartman Martin Chadwick Gaurav Singh Tomar Xavier Garcia Evan Senter Emanuel Taropa Thanumalayan Sankaranarayana Pillai Jacob Devlin Michael Laskin Diego de Las Casas Dasha Valter Connie Tao Lorenzo Blanco Adrià Puigdomènech Badia David Reitter Mianna Chen Jenny Brennan Clara Rivera Sergey Brin Shariq Iqbal Gabriela Surita Jane Labanowski Abhi Rao Stephanie Winkler Emilio Parisotto Yiming Gu Kate Olszewska Yujing Zhang Ravi Addanki Antoine Miech Annie Louis Laurent El Shafey Denis Teplyashin Geoff Brown Elliot Catt Nithya Attaluri Jan Balaguer Jackie Xiang Pidong Wang Zoe Ashwood Anton Briukhov Albert Webson Sanjay Ganapathy Smit Sanghavi Core Contributors Ajay Kannan Ming-Wei Chang Axel Stjerngren Josip Djolonga Yuting Sun Ankur Bapna Matthew Aitchison Pedram Pejman Henryk Michalewski Tianhe Yu Cindy Wang Juliette Love Junwhan Ahn Dawn Bloxwich Kehang Han Peter Humphreys Thibault Sellam James Bradbury Varun Godbole Sina Samangooei Bogdan Damoc Alex Kaskasoli Sébastien M. R. Arnold Vijay Vasudevan Shubham Agrawal Jason Riesa Dmitry Lepikhin Richard Tanburn Srivatsan Srinivasan Hyeontaek Lim Sarah Hodkinson Pranav Shyam Johan Ferret Steven Hand Ankush Garg Tom Le Paine Jian Li Yujia Li Minh Giang Alexander Neitz Zaheer Abbas Sarah York Machel Reid Elizabeth Cole Aakanksha Chowdhery Dipanjan Das Dominika Rogozińska Vitaly Nikolaev Core Contributors Pablo Sprechmann Zachary Nado Lukas Zilka Flavien Prost Luheng He Marianne Monteiro Gaurav Mishra Chris Welty Josh Newlan Dawei Jia Miltiadis Allamanis Clara Huiyi Hu Raoul de Liedekerke Justin Gilmer Carl Saroufim Shruti Rijhwani Shaobo Hou Disha Shrivastava Anirudh Baddepudi Alex Goldin Adnan Ozturel Albin Cassirer Yunhan Xu Daniel Sohn Devendra Sachan Reinald Kim Amplayo Craig Swanson Dessie Petrova Shashi Narayan Arthur Guez Siddhartha Brahma Jessica Landon Miteyan Patel Ruizhe Zhao Kevin Villela Luyu Wang Wenhao Jia Matthew Rahtz Mai Giménez Legg Yeung Hanzhao Lin James Keeling Petko Georgiev Diana Mincu Boxi Wu Salem Haykal Rachel Saputro Kiran Vodrahalli Core Contributors James Qin Zeynep Cankara Abhanshu Sharma Nick Fernando Will Hawkins Behnam Neyshabur Solomon Kim Adrian Hutter Priyanka Agrawal Alex Castro-Ros George van den Driessche Tao Wang Fan Yang Shuo-yiin Chang Paul Komarek Ross McIlroy Mario Lučić Guodong Zhang Wael Farhan Michael Sharman Paul Natsev Paul Michel Yong Cheng Yamini Bansal Siyuan Qiao Kris Cao Siamak Shakeri Christina Butterfield Justin Chung Paul Kishan Rubenstein Shivani Agrawal Arthur Mensch Kedar Soparkar Karel Lenc Timothy Chung Aedan Pope Loren Maggiore Jackie Kay Priya Jhakra Shibo Wang Joshua Maynez Mary Phuong Taylor Tobin Andrea Tacchetti Maja Trebacz Kevin Robinson Yash Katariya Sebastian Riedel Core Contributors Paige Bailey Kefan Xiao Nimesh Ghelani Lora Aroyo Ambrose Slone Neil Houlsby Xuehan Xiong Zhen Yang Elena Gribovskaya Jonas Adler Mateo Wirth Lisa Lee Music Li Thais Kagohara Jay Pavagadhi Sophie Bridgers Anna Bortsova Sanjay Ghemawat Zafarali Ahmed Tianqi Liu Richard Powell Vijay Bolina Mariko Iinuma Polina Zablotskaia James Besley Da-Woon Chung Timothy Dozat Ramona Comanescu Xiance Si Jeremy Greer Guolong Su Martin Polacek Raphaël Lopez Kaufman Simon Tokumine Hexiang Hu Elena Buchatskaya Yingjie Miao Mohamed Elhawaty Aditya Siddhant Nenad Tomasev Jinwei Xing Christina Greer Helen Miller Shereen Ashraf Aurko Roy Zizhao Zhang Ada Ma Angelos Filos Core Contributors Milos Besta Rory Blevins Ted Klimenko Chih-Kuan Yeh Soravit Changpinyo Jiaqi Mu Oscar Chang Mantas Pajarskas Carrie Muir Vered Cohen Charline Le Lan Krishna Haridasan Amit Marathe Steven Hansen Sholto Douglas Rajkumar Samuel Mingqiu Wang Sophia Austin Chang Lan Jiepu Jiang Justin Chiu Jaime Alonso Lorenzo Lars Lowe Sjösund Sébastien Cevey Zach Gleicher Thi Avrahami Anudhyan Boral Hansa Srinivasan Vittorio Selo Rhys May Konstantinos Aisopos Léonard Hussenot Livio Baldini Soares Kate Baumli Michael B. Chang Adrià Recasens Ben Caine Alexander Pritzel Filip Pavetic Fabio Pardo Anita Gergely Justin Frye Vinay Ramasesh Dan Horgan Kartikeya Badola Nora Kassner Subhrajit Roy Ethan Dyer Core Contributors Víctor Campos Alex Tomala Yunhao Tang Dalia El Badawy Elspeth White Basil Mustafa Oran Lang Abhishek Jindal Sharad Vikram Zhitao Gong Sergi Caelles Ross Hemsley Gregory Thornton Fangxiaoyu Feng Wojciech Stokowiec Ce Zheng Phoebe Thacker Çağlar Ünlü Zhishuai Zhang Mohammad Saleh James Svensson Max Bileschi Piyush Patil Ankesh Anand Roman Ring Katerina Tsihlas Arpi Vezer Marco Selvi Toby Shevlane Mikel Rodriguez Tom Kwiatkowski Samira Daruki Keran Rong Allan Dafoe Nicholas FitzGerald Keren Gu-Lemberg Mina Khan Lisa Anne Hendricks Marie Pellat Vladimir Feinberg James Cobon-Kerr Tara Sainath Maribeth Rauh Sayed Hadi Hashemi Richard Ives Yana Hasson YaGuang Li Eric Noland Core Contributors Yuan Cao Nathan Byrd Le Hou Qingze Wang Thibault Sottiaux Michela Paganini Jean-Baptiste Lespiau Alexandre Moufarek Samer Hassan Kaushik Shivakumar Joost van Amersfoort Amol Mandhane Pratik Joshi Anirudh Goyal Matthew Tung Andrew Brock Hannah Sheahan Vedant Misra Cheng Li Nemanja Rakićević Mostafa Dehghani Fangyu Liu Sid Mittal Junhyuk Oh Seb Noury Eren Sezener Fantine Huot Matthew Lamm Nicola De Cao Charlie Chen Contributors Gamaleldin Elsayed Ed Chi Mahdis Mahdieh Ian Tenney Nan Hua Ivan Petrychenko Patrick Kane Dylan Scandinaro Rishub Jain Jonathan Uesato Romina Datta Adam Sadovsky Oskar Bunyan Dominik Rabiej Shimu Wu John Zhang Contributors Gautam Vasudevan Edouard Leurent Mahmoud Alnahlawi Ionut Georgescu Nan Wei Ivy Zheng Betty Chan Pam G Rabinovitch Piotr Stanczyk Ye Zhang David Steiner Subhajit Naskar Michael Azzam Matthew Johnson Adam Paszke Chung-Cheng Chiu Jaume Sanchez Elias Afroz Mohiuddin Faizan Muhammad Jin Miao Andrew Lee Nino Vieillard Sahitya Potluri Jane Park Elnaz Davoodi Jiageng Zhang Jeff Stanway Drew Garmon Abhijit Karmarkar Zhe Dong Jong Lee Aviral Kumar Luowei Zhou Jonathan Evens William Isaac Zhe Chen Johnson Jia Anselm Levskaya Zhenkai Zhu Chris Gorgolewski Peter Grabowski Yu Mao Alberto Magni Kaisheng Yao Javier Snaider Norman Casagrande Paul Suganthan Evan Palmer Contributors Geoffrey Irving Edward Loper Manaal Faruqui Isha Arkatkar Nanxin Chen Izhak Shafran Michael Fink Alfonso Castaño Irene Giannoumis Wooyeol Kim Mikołaj Rybiński Ashwin Sreevatsa Jennifer Prendki David Soergel Adrian Goedeckemeyer Willi Gierke Mohsen Jafari Meenu Gaba Jeremy Wiesner Diana Gage Wright Yawen Wei Harsha Vashisht Yana Kulizhskaya Jay Hoover Maigo Le Lu Li Chimezie Iwuanyanwu Lu Liu Kevin Ramirez Andrey Khorlin Albert Cui Tian LIN Marin Georgiev Marcus Wu Ricardo Aguilar Keith Pallo Abhishek Chakladar Alena Repina Xihui Wu Tom van der Weide Priya Ponnapalli Caroline Kaplan Jiri Simsa Shuangfeng Li Olivier Dousse Fan Yang Jeff Piper Nathan Ie Contributors Minnie Lui Rama Pasumarthi Nathan Lintz Anitha Vijayakumar Lam Nguyen Thiet Daniel Andor Pedro Valenzuela Cosmin Paduraru Daiyi Peng Katherine Lee Shuyuan Zhang Somer Greene Duc Dung Nguyen Paula Kurylowicz Sarmishta Velury Sebastian Krause Cassidy Hardin Lucas Dixon Lili Janzer Kiam Choo Ziqiang Feng Biao Zhang Achintya Singhal Tejasi Latkar Mingyang Zhang Quoc Le Elena Allica Abellan Dayou Du Dan McKinnon Natasha Antropova Tolga Bolukbasi Orgad Keller David Reid Daniel Finchelstein Maria Abi Raad Remi Crocker Peter Hawkins Robert Dadashi Colin Gaffney Sid Lall Ken Franko Egor Filonov Anna Bulanova Rémi Leblond Vikas Yadav Shirley Chung Harry Askham Luis C. Cobo Contributors Kelvin Xu Felix Fischer Jun Xu Christina Sorokin Chris Alberti Chu-Cheng Lin Colin Evans Hao Zhou Alek Dimitriev Hannah Forbes Dylan Banarse Zora Tung Jeremiah Liu Mark Omernick Colton Bishop Chintu Kumar Rachel Sterneck Ryan Foley Rohan Jain Swaroop Mishra Jiawei Xia Taylor Bos Geoffrey Cideron Ehsan Amid Francesco Piccinno Xingyu Wang Praseem Banzal Petru Gurita Hila Noga Premal Shah Daniel J. Mankowitz Alex Polozov Nate Kushman Victoria Krakovna Sasha Brown MohammadHossein Bateni Dennis Duan Vlad Firoiu Meghana Thotakuri Tom Natan Anhad Mohananey Matthieu Geist Sidharth Mudgal Sertan Girgin Hui Li Jiayu Ye Ofir Roval Reiko Tojo Contributors Michael Kwong James Lee-Thorp Christopher Yew Quan Yuan Sumit Bagri Danila Sinopalnikov Sabela Ramos John Mellor Abhishek Sharma Aliaksei Severyn Jonathan Lai Kathy Wu Heng-Tze Cheng David Miller Nicolas Sonnerat Denis Vnukov Rory Greig Jennifer Beattie Emily Caveness Libin Bai Julian Eisenschlos Alex Korchemniy Tomy Tsai Mimi Jasarevic Weize Kong Phuong Dao Zeyu Zheng Frederick Liu Fan Yang Rui Zhu Mark Geller Tian Huey Teh Jason Sanmiya Evgeny Gladchenko Nejc Trdin Andrei Sozanschi Daniel Toyama Evan Rosen Sasan Tavakkol Linting Xue Chen Elkind Oliver Woodman John Carpenter George Papamakarios Rupert Kemp Sushant Kafle Tanya Grunina Rishika Sinha Contributors Alice Talbert Abhimanyu Goyal Diane Wu Denese Owusu-Afriyie Cosmo Du Chloe Thornton Jordi Pont-Tuset Pradyumna Narayana Jing Li Saaber Fatehi John Wieting Omar Ajmeri Benigno Uria Tao Zhu Yeongil Ko Laura Knight Amélie Héliou Ning Niu Shane Gu Chenxi Pang Dustin Tran Yeqing Li Nir Levine Ariel Stolovich Norbert Kalb Rebeca Santamaria-Fernandez Sonam Goenka Wenny Yustalim Robin Strudel Ali Elqursh Balaji Lakshminarayanan Charlie Deck Shyam Upadhyay Hyo Lee Mike Dusenberry Zonglin Li Xuezhi Wang Kyle Levin Raphael Hoffmann Dan Holtmann-Rice Olivier Bachem Summer Yue Sho Arora Eric Malmi Daniil Mirylenka Qijun Tan Christy Koh Soheil Hassas Yeganeh Contributors Siim Põder Steven Zheng Francesco Pongetti Mukarram Tariq Yanhua Sun Lucian Ionita Mojtaba Seyedhosseini Pouya Tafti Ragha Kotikalapudi Zhiyu Liu Anmol Gulati Jasmine Liu Xinyu Ye Bart Chrzaszcz Lily Wang Nikhil Sethi Tianrun Li Ben Brown Shreya Singh Wei Fan Aaron Parisi Joe Stanton Chenkai Kuang Vinod Koverkathu Christopher A. Choquette-Choo Yunjie Li TJ Lu Abe Ittycheriah Prakash Shroff Pei Sun Mani Varadarajan Sanaz Bahargam Rob Willoughby David Gaddy Ishita Dasgupta Guillaume Desjardins Marco Cornero Brona Robenek Bhavishya Mittal Ben Albrecht Ashish Shenoy Fedor Moiseev Henrik Jacobsson Alireza Ghaffarkhah Morgane Rivière Alanna Walton Clément Crepy Alicia Parrish Contributors Yuan Liu Zongwei Zhou Clement Farabet Carey Radebaugh Praveen Srinivasan Claudia van der Salm Andreas Fidjeland Salvatore Scellato Eri Latorre-Chimoto Hanna Klimczak-Plucińska David Bridson Dario de Cesare Tom Hudson Piermaria Mendolicchio Lexi Walker Alex Morris Ivo Penchev Matthew Mauger Alexey Guseynov Alison Reid Seth Odoom Lucia Loher Victor Cotruta Madhavi Yenugula Dominik Grewe Anastasia Petrushkina Tom Duerig Antonio Sanchez Steve Yadlowsky Amy Shen Amir Globerson Adam Kurzrok Lynette Webb Sahil Dua Dong Li Preethi Lahoti Surya Bhupatiraju Dan Hurt Haroon Qureshi Ananth Agarwal Tomer Shani Matan Eyal Anuj Khare Shreyas Rammohan Belle Lei Wang Chetan Tekur Mihir Sanjay Kale Jinliang Wei Contributors Ruoxin Sang Brennan Saeta Tyler Liechty Yi Sun Yao Zhao Stephan Lee Pandu Nayak Doug Fritz Manish Reddy Vuyyuru John Aslanides Nidhi Vyas Martin Wicke Xiao Ma Taylan Bilal Evgenii Eltyshev Daniel Balle Nina Martin Hardie Cate James Manyika Keyvan Amiri Yelin Kim Xi Xiong Kai Kang Florian Luisier Nilesh Tripuraneni David Madras Mandy Guo Austin Waters Oliver Wang Joshua Ainslie Jason Baldridge Han Zhang Garima Pruthi Jakob Bauer Feng Yang Riham Mansour Jason Gelman Yang Xu George Polovets Ji Liu Honglong Cai Warren Chen XiangHai Sheng Emily Xue Sherjil Ozair Adams Yu Christof Angermueller Contributors Xiaowei Li Weiren Wang Julia Wiesinger Emmanouil Koukoumidis Yuan Tian Anand Iyer Madhu Gurumurthy Mark Goldenson Parashar Shah MK Blake Hongkun Yu Anthony Urbanowicz Jennimaria Palomaki Chrisantha Fernando Kevin Brooks Ken Durden Harsh Mehta Nikola Momchev Elahe Rahimtoroghi Maria Georgaki Amit Raul Sebastian Ruder Morgan Redshaw Jinhyuk Lee Komal Jalan Dinghua Li Ginger Perng Blake Hechtman Parker Schuh Milad Nasr Mia Chen Kieran Milan Vladimir Mikulik Trevor Strohman Juliana Franco Tim Green Program Leads Demis Hassabis Koray Kavukcuoglu Overall Technical Leads (equal contribution) Jeffrey Dean Oriol Vinyals The roles are defined as below: • : Individual(s) responsible for the sub-team throughout the project. Lead Individual that had significant impact throughout the project. • Core Contributor: : Individual that had contributions to the project and was partially involved with the effort. • Contributor Responsible for the organizational aspects of the Gemini effort • Program Lead: Responsible for the technical direction of the overall Gemini effort • Overall Technical Lead: Within each role, contributions are equal, and are listed in a randomized order. Ordering within each role does not indicate ordering of the contributions. Gemini is a cross-Google effort, with members from Google DeepMind (GDM), Google Research (GR), Knowledge and Information (K&I), Core ML, Cloud, Labs, and more. We thank our reviewers and colleagues for their valuable discussions and feedback on the report— Alexandra Belias, Arielle Bier, Eleanor Tomlinson, Emily Hossellman, Gaby Pearl, Helen King, Hollie Dobson, Jaclyn Konzelmann, Jennifer Beroshi, Joel Moss, Jon Small, Jonathan Fildes, Oli Gaymond, Rebecca Bland, Reena Jana, and Tom Lue. Our work is made possible by the dedication and efforts of numerous teams at Google. We would like to acknowledge the support from Abhi Mohan, Adekunle Bello, Aishwarya Nagarajan, Alaa Saade, Alejandro Lince, Alexander Chen, Alexander Kolbasov, Alexander Schiffhauer, Amar Subramanya, Ameya Shringi, Amin Vahdat, Anda Rabatić, Anthonie Gross, Antoine Yang, Anthony Green, Anton Ruddock, Art Khurshudov, Artemis Chen, Arthur Argenson, Avinatan Hassidim, Beiye Liu, Benjamin Schroeder, Bin Ni, Brett Daw, Bryan Chiang, Burak Gokturk, Carl Crous, Carrie Grimes Bostock, Charbel Kaed, Charlotte Banks, Che Diaz, Chris Larkin, Christy Lian, Claire Cui, Clare Bycroft, Corentin Tallec, Daniel Herndon, Dave Burke, David Battle, David Engel, Dipannita Shaw, Donghyun Koo, Doug Ritchie, Dragos Stefanescu, Elissa Wolf, Emre Sargin, Eran Ofek, Eric Herren, Estella King, Fatema Alkhanaizi, Felix Gimeno, Fernando Pereira, Florent Altché, Gabriel Carvajal, Gaurav Gandhi, George Powell, Goran Pavičić, Harry Richardson, Hassan Wassel, Hongji Li, Idan Szpektor, Igor Ivanisevic, Ivan Jambrešić, Ivan Jurin, Jade Fowler, James Assiene, Jay Yagnik, Jeanbastien Grill, Jeff Seibert, Jenna LaPlante, Jessica Austin, Jianxing Lu, Jim O’Keeffe, Jin Huang, Joe Heyward, Johannes Welbl, John Jumper, Jonathan Caton, Josh Woodward, Joshua Foster, Kathryn Tunyasuvunakool, Katrina Wong, Kavya Kopparapu, Kelvin Nguyen, Kira Yin, Konstantin Sharlaimov, Kun Li, Lee Hong, Lilly Taylor, Longfei Shen, Luc Mercier, Maciej Mikuła, Mania Abdi, Manuel Sanchez, Maria Ines Aranguren, Mario Carlos Cortes III, Matthew Tait, Matthias Lochbrunner, Mehdi Ghissassi, Micah Mosley, Michael Bendersky, Michael Figurnov, Michael Harris, Michael Mathieu, Michael O’Neill, Michael Vorburger, Mihir Paradkar, Nandita Dukkipati, Nathan Carter, Nathan Watson, Neil Rabinowitz, Nikhil Dandekar, Nishant Ranka, Obaid Sarvana, Olcan Sercinoglu, Olivier Lacombe, Ottavia Bertolli, Paul Caron, Pranesh Srinivasan, Praveen Kumar, Rahul Sukthankar, Raia Hadsell, Rajagopal Ananthanarayanan, Roberto Lupi, Rosie Zou, Sachin Menezes, Sadegh Jazayeri, Sam Cheung, Sameer Bidichandani, Sania Alex, Sanjiv Kumar, Sara Wiltberger, Sarah Fitzgerald, Saz Basu, Sebastian Nowozin, Shannon Hepburn, Shayne Cardwell, Sissie Hsiao, Srinivasan Venkatachary, Sugato Basu, Sundar Pichai, Sundeep Tirumalareddy, Susannah Young, Swetha Vijayaraghavan, Tania Bedrax-Weiss, Taylor Applebaum, Teiva Harsanyi, Terry Chen, Tim Blyth, Ting Liu, Tom Cobley, Tomas Izo, Trystan Upstill, Varun Singhai, Vedrana Klarić Trupčević, Victor Cai, Vladimir Pudovkin, Vu Dang, Wenbo Zhao, Wesley Crow, Wesley Szeng, Xiaodan Song, Yazhou Zu, Ye Tian, Yicong Wang, Yixing Wang, Yossi Matias, Yunlong Jiao, Zachary Jessup, Zhenchuan Pang, Žiga Avsec, Zimeng Yang, and Zoubin Ghahramani. We’d also like to recognize the AlphaCode team, the Borg Scheduling team, the Facilities team, the Gemini Demo Team, the Global Server Ops (GSO) team, the JAX team, the the Legal team, ML SRE team, the ML Supercomputer (MLSC) team, the PartIR team, the Platforms Infrastructure Engineering (PIE) team, and the XLA Compiler team. We thank everyone at Google not explicitly mentioned above, who have shared excitement, given feedback on early Gemini models or created interesting demo uses of Gemini, and worked with or supported the core Gemini team on many aspects of this project.

Gemini -  A Family of Highly Capable Multimodal Models: Contributions and Acknowledgments

The United States v Meta Platforms Court Filing October 24, 2023 is part of . You can jump to any part in this filing . This is part 17 of 100. HackerNoon’s Legal PDF Series here 3. Meta’s Recommendation Algorithms encourage compulsive use, which Meta does not disclose. 151. Instagram and Facebook employ Recommendation Algorithms that curate content from the main feeds and other parts of the Platforms. 152. The Recommendation Algorithms use data points, or “signals,” harvested from individual users to choose and/or arrange each new piece of content to display to a user. Such signals include, but are not limited to, overt actions like Liking a post or following a page as well as such unconscious actions such as lingering on—but not otherwise engaging with—certain content or visiting but not following another user’s page. 153. Meta employs Recommendation Algorithms universally across its Social Media Platforms, including the Instagram Platform’s Main Feed (the scrolling presentation of content immediately visible upon opening the app) and Explore Feed (another scrolling presentation of algorithmically curated content that can be guided by a user’s text input in a search field). 154. Meta designed its Recommendation Algorithms to maximize youth engagement in several ways but did not disclose these engagement-maximization features to the public—instead representing that these algorithms were intended to benefit the user. 155. First, Meta designed the Recommendation Algorithms to present material to young users in an unpredictable sequence rather than displaying posts chronologically. 156. Specifically, Meta’s Recommendation Algorithms display content to young users through a sequencing method referred to by psychologists as “variable reinforcement schedules” or “variable reward schedules.” 157. As Dr. Mark D. Griffiths, Distinguished Professor of Behavioral Addiction at Nottingham Trent University, explains: The rewards [experienced on social media platforms]—which may be physiological, psychological and/or social—can be infrequent but even the anticipation of one of these rewards can be psychologically and/or physiologically pleasing. The rewards are what psychologists refer to as variable reinforcement schedules and is one of the main reasons why social media users repeatedly check their screens. Social media sites are ‘chock-ablock’ with unpredictable rewards. Habitual social media users never know if their next message or notification will be the one that makes them feel really good. In short, random rewards keep individuals responding for longer and has been found in other activities such as the playing of slot machines and video games.[9] 158. Because they do not work in a predictable pattern, these “variable reinforcement schedules” trigger a release of dopamine, a neurotransmitter released by the brain in response to certain stimuli. Dopamine, commonly “seen to be the ‘pleasure chemical,’” is released in anticipation of a potential reward. However, dopamine neurons fire for only a relatively short period of time, and after dopamine is released, an “individual can become disheartened and disengaged.”[10 ] 159. As researchers Rasan Burhan and Jalal Moradzadeh explain, the variable reinforcement schedules baked into social media platforms like Instagram can lead to “addiction with dopamine implicated”: [T]he user can be kept in a loop. Essentially, that’s how the social media apps exploit these innate systems. The way this comes about is through a term referred to as Variable Reward Schedules. This works by positive stimuli being provided at random intervals. By users checking their phones for notifications and updates at periodic intervals for something that could be intrinsically rewarding. Most of the time it’s a neutral stimuli, but on occasion there may be a positive stimuli leading to the rewarding dopamine release hence keeping the user in the feedback loop.[11] 160. [Redacted] 161. By algorithmically serving content to young users according to variable reward schedules, Meta manipulates dopamine releases in its young users, inducing them to engage repeatedly with its Platforms—much like a gambler at a slot machine. 161-164. [Redacted] 165. Nonetheless, and as illustrated above, as recently as 2020, Meta continued to intentionally design its Platforms to manipulate dopamine responses in its young users to maximize time spent on its Platforms. Meta did not disclose that its algorithms were designed to capitalize on young users’ dopamine responses and create an addictive cycle of engagement. 166. Second, Meta uses data harvested from its users to target user engagement on an individual level via its Recommendation Algorithms—making continued engagement even more difficult for young users to resist. 167. In a June 8, 2021 public blog post on Instagram’s website, Mosseri stated that Meta collects and supplies its Recommendation Algorithms with thousands of “signals” across Instagram’s Feed and Stories, including “[y]our activity” and “[y]our history of interacting with someone.” Mosseri’s post explained that the collection of “[y]our activity . . . helps us understand what you might be interested in . . .” and the collection of “[y]our history of interacting with someone . . . gives us a sense of how interested you are generally in seeing posts from a particular person.” 168. Similarly, Facebook’s Vice President of Global Affairs wrote in Medium on March 31, 2021, about Facebook’s Recommendation Algorithms: “The goal is to make sure you see what you find most meaningful—not to keep you glued to your smartphone for hours on end. You can think about this sort of like a spam filter in your inbox: it helps filter out content you won’t find meaningful or relevant, and prioritizes content you will.” 169. Likewise, Meta’s terms of service on data collection state that Meta uses user data to “[p]rovide, personalize and improve our Products,” “[p]rovide measurement, analytics, and other business services,” “[p]romote safety, integrity and security,” “[c]ommunicate with you,” and “[r]esearch and innovate for social good.” 170. In reality, though, Meta tracks and logs the behavior of millions of young users and utilizes that data to refine and strengthen the features that induce young users’ compulsive Social Media Platform use. 171. As young users engage with Meta’s Social Media Platforms, they are unwittingly training Meta’s Recommendation Algorithms to provide the particular flow of content, notifications, and features that will most effectively keep them online. 172. Again, Meta does not disclose to consumers that it is weaponizing young users’ data to capture and keep their attention. 173. Meta admits in its Privacy Policy that it uses data provided by its young users for purposes other than facilitating meaningful social experiences, such as “improv[ing] our Products . . . includ[ing] personalizing features, content and recommendations, such as your Facebook Feed, Instagram feed, Stories, and ads.” 174. This includes using young users’ data to “[t]est out new products and features to see if they work” and to “[g]et feedback on our ideas for products or features.” 175. But Meta’s representations about its Recommendation Algorithms do not effectively apprise young users of the reality that Meta is harvesting vast amounts of personal data to train its Recommendation Algorithms to induce them to keep using the Platforms. 176. Third, the Recommendation Algorithms increase young users’ engagement by periodically presenting those users with psychologically and emotionally gripping content, including content related to eating disorders, violent content, content encouraging negative selfperception and body image issues, bullying content, and other categories of content known by Meta to provoke intense reactions. 177. Meta’s Recommendation Algorithms are optimized to promote user engagement. Serving harmful or disturbing content has been shown to keep young users on the Platforms longer. Accordingly, the Recommendation Algorithms predictably and routinely present young users with psychologically and emotionally distressing content that induces them to spend increased time on the Social Media Platforms. And, once a user has interacted with such harmful content, the Recommendation Algorithm feeds that user additional similar content. 178-182. [Redacted] 183. Again, though, Meta’s public statements regarding its algorithms’ amplification of distressing and problematic content did not reflect Meta’s true awareness of these problems. 184. In fact, Meta has strongly denied that its Social Media Platforms amplify extreme, distressing, or problematic content. 185. For example, on September 30, 2021, Davis denied that Meta promotes harmful content, such as content promoting eating disorders to youth, when she testified before Congress, stating, “we do not direct people towards content that promotes eating disorders. That actually violates our policies, and we remove that content when we become aware of it. We actually use AI to find content like that and remove it.” 186. [Redacted] 187. Likewise, in a June 8, 2021 post on the Instagram website, titled “Shedding More Light on How Instagram Works,” Mosseri describes Meta’s Recommendation Algorithms by providing examples of benign content recommendations (e.g., “if you’re interested in dumplings you might see posts about related topics, like gyoza and dim sum . . .”). The post provides no accompanying examples or warnings disclosing that the Recommendation Algorithms also tend to suggest content that is dangerous or harmful for young users. 188. The Instagram website also boasts that “[a]t Instagram, we have guidelines that govern what content we recommend to people” and specifies that Instagram “avoid[s] making recommendations that may be inappropriate for younger viewers . . . . We use technology to detect both content and accounts that don’t meet these Recommendations Guidelines and to help us avoid recommending them. As always, content that goes against our Community Guidelines will be removed from Instagram.” 189. A parent or young user encountering these and similar communications by Meta could reasonably understand Meta to be representing that its Recommendation Algorithms do not promote content to young users that violates Meta’s Recommendation Guidelines or is otherwise dangerous or inappropriate for young users. 190. But as explained above, Meta does increase young users’ engagement with its Platforms by periodically presenting them with psychologically and emotionally gripping content, including content related to eating disorders, violent content, content encouraging negative self perception and body image issues, bullying content, and other categories of content known by Meta to provoke intense reactions from users. [9] Mark D. Griffiths, Adolescent Social Networking: How Do Social Media Operators Facilitate Habitual Use?, 36 Educ. & Health J. 66, 67 (2018), (internal references omitted). http://archive.today/cPgJ1 [10] Rasan Burhan & Jalal Moradzadeh, Neurotransmitter Dopamine (DA) and its Role in the Development of Social Media Addiction, 11 J. Neurology & Neurophysiology 1, 1 (2020), . http://archive.today/kxldL [11] Id. at 1-2. Continue Reading . Here About HackerNoon Legal PDF Series: We bring you the most important technical and insightful public domain court case filings. This court case 4:23-cv-05448 retrieved on October 25, 2023, from is part of the public domain. The court-created documents are works of the federal government, and under copyright law, are automatically placed in the public domain and may be shared without legal restriction. Washingtonpost.com

Facebook and Instagram Use Algorithms That Promote Compulsive Use

AI has been slaying it in almost every sector. Healthcare is benefitting a lot from voice based . But this doesn’t mean we can trust them blindly. LLMs and ML Google DeepMind an AI program named AlphaMissense, which predicts the harm a missense mutation can do, when a single DNA letter is altered. The tool, which leverages data from 71 million single-letter mutations affecting human proteins, uses machine learning to “understand” the language of proteins, helping it score the risk level of genetic changes. created What AI is doing in synthetic biology is nothing short of revolutionary. Using AI, scientists are to create new unseen proteins like coding software. We can expect groundbreaking changes in not only medicine, but also in sectors like agriculture, energy, and climate. now programming cells Also, recently, researchers from Edith Cowan University software that quickly scans bone density scans to detect abdominal aortic calcification (AAC), which is a predictor of cardiovascular events and other health risks. The software processed images with 80% accuracy and could revolutionize early disease detection during routine clinical practice. developed In August, a ML model to identify autism spectrum disorder with 95% accuracy. While Stanford Medicine developed an algorithm that reviews images to determine a brain tumour’s aggressiveness, its genetic makeup, and effectiveness of surgery. “It’s sort of a decision support system for the physicians,” one researcher says. international researchers trained Also, Scientists at the Osaka Metropolitan University an AI model that accurately identifies cardiac functions and valvular heart diseases using chest radiographs. The research could supplement traditional echocardiography, improve diagnostic efficiency, and be especially useful in settings lacking specialized technicians. developed Startups in are also upping their game with AI. London-based Ultromics its AI to detect a heart condition that human eyes can often not detect. In July, physician-entrepreneurs at Israel’s started training AI to diagnose cardiac disease on a simple tablet, where the doctors don’t have to be heart experts. With a handheld scanner and tablet, non-cardiologists get a full diagnostic on 8 key heart readings. With its training data, the AI platform already has more experience than any living cardiologist, the physicians say. HealthTech trained Sheba Medical Center In September, , an Indian AI-powered grocery scoring and recommendation platform, tied up with, (Microbiome Research Pvt Ltd), a research platform in the field of gut microbiome and, , a diagnostics startup. With a shared commitment to preventive healthcare, the trio seeks to empower individuals with individualised in-depth diagnostics and convert that information into actionable real-time intelligence. Xume Microbiome.in TruDiagno This can help individuals to make informed choices based on the state of one’s gut microbiome and key diagnostic markers. Microbiome.in is the only private lab with an in-house Illumina and Nanopore (portable) Sequencer enabling unparalleled research on the role of gut microbiome, while TruDiagno adds advanced molecular diagnostic capabilities to the mix. Imagine the power of understanding our microbiome and diagnostic markers to drive actionable intelligence; where the answer is not giving up our favourite foods but simply making better choices based on what our gut and body are telling us Akshaye Jalan, Founder & CEO, Xume “We’re all unique, so how can a generic solution mired in restriction be the answer to good health? Imagine the power of understanding our microbiome and diagnostic markers to drive actionable intelligence; where the answer is not giving up our favourite foods but simply making better choices based on what our gut and body are telling us. This is the future of health and wellness and what our partnership with Microbiome.in and TruDiagno endeavours to do.” said Akshaye Jalan, Founder & CEO, Xume. In September, Oracle introduced voice-activated AI to hospitals, which will help doctors and clinicians to automate medical note-taking, order medications and view lab reviews while also more easily reviewing a patient’s medical record. This is not about going from 10 clicks to seven clicks or six clicks to five clicks. This is about completely eliminating clicks, where we can and where it makes sense. That is not a tomorrow or that’s years away. That is a tomorrow that’s here Mike Sicilia, executive vice president, industries at Oracle “We’ve completely rethought how long it takes somebody to get their job done using the system. I think that we are on the verge of completely eliminating ‘pajama time’ as a result. This is not about going from 10 clicks to seven clicks or six clicks to five clicks. This is about completely eliminating clicks, where we can and where it makes sense. That is not a tomorrow or that’s years away. That is a tomorrow that’s here,” Mike Sicilia, executive vice president, industries at Oracle told the audience at the . Oracle Health Conference But does this mean AI is your next doctor instead of a human? All the above instances are brilliant examples of AI helping humankind become better at healthcare, but they’re also being done under supervision of experts, who are human. The general online population has been turning to Dr. Google even before the Red Hot Chilli Peppers wrote . But there’s a new doctor in town, Dr. ChatGPT. , “the most recent version of ChatGPT made zero “grossly inaccurate” statements when presented with a standard set of eye complaints.” Californication According to NPR However, questions still remain about how to fit in LLMs into healthcare with similar safeguards that are applied when introducing new drugs or medical devices. This article was originally published by in Navanwita Bora Sachdev The Tech Panda.

Dr. ChatGPT vs. Dr. Google: An Assessment of AI Medical Advice

At length Louis XIV, bowed beneath the weight of a reign of sixty years

The field of natural language processing (NLP) has seen a significant shift since the introduction of transformer architectures. First proposed in the "Attention is All You Need" paper [5], these architectures have formed the basis for a variety of powerful models, including and BERT GPT. In this article, we will overview transformer architectures. We will dissect the encoder-only architecture of , exploring its masked language modeling and next-sentence prediction tasks. Next, we cover the decoder-only architecture of GPT models, highlighting its use of causal self-attention and scalability, we will also discuss why this architecture became dominant for general-purpose language models. Additionally, we will revisit the original Transformer architecture, which consists of both an encoder and a decoder. BERT BERT - Bidirectional Encoder Representations from Transformers [1] Overview BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained transformer-based model developed by researchers at AI Language. BERT revolutionized the NLP landscape by introducing a powerful encoder-only architecture that leverages bidirectional context for a wide range of tasks. The primary innovation of BERT lies in its ability to generate context-specific word embeddings, allowing it to capture complex language patterns and dependencies. This understanding of context has several important implications, which we will explore in more detail below. Google Training details The BERT model is trained on two tasks: Masked Language Modeling, where the objective is to predict the masked word based on the context (Fig. 1), and Next Sentence Prediction, where the objective is to classify whether sentence B follows Sentence A in the text. Context-Dependent Word Embeddings Traditional word embedding techniques, such as Word2Vec and GloVe, represent words as fixed-dimensional vectors that capture their general meaning. However, these embeddings do not account for the context in which words appear, which can limit their effectiveness in the modeling of language. BERT addresses this limitation by generating word embeddings that depend on the context in which the words are used. This enables BERT to capture the nuances of language and understand words with multiple meanings based on their surrounding tokens. Understanding word in the context -> Understanding the context One key insight of BERT is that the ability to generate context-dependent word embedding implies that the model understands the context itself. By learning to predict masked tokens based on their surrounding words, BERT demonstrates its ability to infer word meanings from context. This suggests that the model has developed a strong grasp of the language structure and relationships between words, which is essential for effective transfer learning. Transfer Learning and Task Adaptation BERT's understanding of context makes it particularly powerful for transfer learning, as its pre-trained representations can be easily fine-tuned for various downstream tasks. This adaptability is a key strength of BERT, as it allows the model to achieve state-of-the-art performance on a wide range of NLP tasks with minimal architectural changes and relatively small amounts of task-specific data. Examples of tasks for which BERT has been successfully fine-tuned include: Sentiment analysis: BERT can be used to classify text based on sentiment, such as determining whether a movie review is positive or negative. Named entity recognition: Fine-tuning BERT can enable it to identify and classify entities, such as people, organizations, and locations, in a given text. Summary In summary, BERT's encoder-only architecture, context-dependent word embeddings, and deep understanding of context have made it a powerful and versatile model for transfer learning. Its success in a wide variety of NLP tasks has demonstrated the potential of transformer-based models and paved the way for subsequent developments in the field. GPT - Generative Pre-trained Transformer Overview GPT (Generative Pre-trained Transformer) is a state-of-the-art transformer-based model developed by OpenAI. The third iteration in the GPT series, GPT-3 builds upon the success of its predecessors by scaling up the architecture and leveraging a massive amount of pre-training data. In contrast to BERT, GPT is based on the decoder-only Transformer architecture (Fig. 2). This section will explore the core insights of GPT, and highlight its important features that contribute to its success while drawing parallels with BERT. Language Understanding through Next-Word Prediction While BERT learns context-specific word embeddings through its masked language modeling task, GPT-3 focuses on predicting the next word in a given sequence, given the preceding words. Similar to BERT, this word prediction task allows GPT to develop a deep understanding of the language, as it must learn syntactic, semantic, and pragmatic aspects of the text in order to accurately predict what comes next. Efficient Training with Causal Self-Attention The first thing that contributes to the success of GPT models is the efficiency of the training process. GPT employs causal self-attention, a mechanism that allows tokens to only attend to preceding tokens during training. This design choice ensures that the model does not have access to future tokens when predicting the next word, making the training process more efficient. By processing a single text sample of length n, GPT effectively generates n training samples, as it predicts each token based on its preceding context. This increased efficiency in training enables GPT to scale up to larger architectures and datasets, resulting in improved performance and language understanding. In contrast, BERT uses masked language modeling, which involves masking a fixed subset of words and predicting them based on bidirectional context, requiring more computational resources during training. Zero-Shot and Few-Shot Learning The second thing that makes GPT models so interesting is the emergent ability to perform Zero-/ few-shot learning, i.e., by simply providing the model with a task description and zero or few examples of the desired output, GPT-3 can adapt to a wide range of NLP tasks without any task-specific fine-tuning (left image on Fig. 3). GPT-3 develops this remarkable capability only when it gets large enough, i.e., 100Bn+ parameters (right image on Fig.3). It’s possible to train GPT models of this size only due to the training efficiency described in Sec. 3.3, thus, it’s unlikely that BERT-like models will ever demonstrate few-shot learning capabilities and they would always require fine-tuning in practice. The transfer learning capability of GPT-3 is a testament to the model's robust language understanding and its ability to generalize from the vast amount of pre-training data. Conclusion In conclusion, 's success in the field of natural language processing can be attributed to its efficient training based on the next word prediction, and zero-shot/few-shot learning capabilities. Its scalability allows it to handle larger architectures and datasets, enhancing its performance and language understanding. GPT-3 Unlike BERT which requires task-specific fine-tuning and more computational resources during training, GPT-3 can adapt to various NLP tasks with minimal task-specific adjustments, highlighting its capacity for generalization from pre-training data. Encoder-Decoder Architecture in Transformers For completeness here we want to overview the original encoder-decoder architecture (Fig. 4) proposed in the seminal "Attention is All You Need" paper [5]. This architecture is used in practice for a wide range of NLP tasks. We already covered the encoder and decoder architectures, the encoder-decoder structure blends those two parts: the encoder processes the input sequence, and the decoder generates the output sequence, leveraging the representations from the encoder. Here we want to highlight a significant difference between encoder-decoder Transformer models with BERT and GPT. Encoder-decoder models are typically specialized for specific tasks, such as translation or text summarization. This means a single model is typically trained to perform one specific function. On the other hand, BERT and GPT are designed to perform a wide array of tasks, and they excel at transfer learning. This difference underscores the versatility of BERT and GPT, while also highlighting the focused strength of task-specific encoder-decoder models. References: [1] 1810.04805.pdf (arxiv.org) [2] The Illustrated BERT, ELMo, and co. (How NLP Cracked Transfer Learning) – Jay Alammar – Visualizing machine learning one concept at a time. (jalammar.github.io) [3] Transformer Decoder-Only 模型批量生成 Trick - 知乎 (zhihu.com) [4] 2005.14165.pdf (arxiv.org) [5] 1706.03762.pdf (arxiv.org) The lead image for this article was generated by HackerNoon's AI Image Generator via the prompt "Illustrate two humanoid robots squaring off"

JALA

jala.tech

-- employees

Since n.d.

COMPANY RANKING

JALA

EVERGREEN INDEX #5723

Representing this company? Click here to claim and customize this page!

Jala WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

JALA

jala.tech

-- employees

Since n.d.

COMPANY RANKING

JALA

EVERGREEN INDEX #5723

RELATED COMPANIESRANK

STORIES ABOUT JALA

Representing this company? Click here to claim and customize this page!

Jala WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!