FedBlock: A Blockchain Approach to Federated Learning against Backdoor Attacks

Autor: Nguyen, Duong H., Nguyen, Phi L., Nguyen, Truong T., Pham, Hieu H., Tran, Duc A.

Federated Learning (FL) is a machine learning method for training with private data locally stored in distributed machines without gathering them into one place for central learning. Despite its promises, FL is prone to critical security risks. First

Externí odkaz: http://arxiv.org/abs/2411.02773

Vintern-1B: An Efficient Multimodal Large Language Model for Vietnamese

Autor: Doan, Khang T., Huynh, Bao G., Hoang, Dung T., Pham, Thuc D., Pham, Nhat H., Nguyen, Quan T. M., Vo, Bang Q., Hoang, Suong N.

In this report, we introduce Vintern-1B, a reliable 1-billion-parameters multimodal large language model (MLLM) for Vietnamese language tasks. By integrating the Qwen2-0.5B-Instruct language model with the InternViT-300M-448px visual model, Vintern-1

Externí odkaz: http://arxiv.org/abs/2408.12480

The Belle II Detector Upgrades Framework Conceptual Design Report

Autor: Aihara, H., Aloisio, A., Auguste, D. P., Aversano, M., Babeluk, M., Bahinipati, S., Banerjee, Sw., Barbero, M., Baudot, J., Beaubien, A., Becherer, F., Bergauer, T., Bernlochner., F. U., Bertacchi, V., Bertolone, G., Bespin, C., Bessner, M., Bettarini, S., Bevan, A. J., Bhuyan, B., Bona, M., Bonis, J. F., Borah, J., Bosi, F., Boudagga, R., Bozek, A., Bračko, M., Branchini, P., Breugnon, P., Browder, T. E., Buch, Y., Budano, A., Campajola, M., Casarosa, G., Cecchi, C., Chen, C., Choudhury, S., Corona, L., de Marino, G., De Nardo, G., De Pietro, G., de Sangro, R., Dey, S., Dingfelder, J. C., Dong, T. V., Dorokhov, A., Dujany, G., Epifanov, D., Federici, L., Ferber, T., Fillinger, T., Finck, Ch., Finocchiaro, G., Forti, F., Frey, A., Friedl, M., Gabrielli, A., Gaioni, L., Gao, Y., Gaudino, G., Gaur, V., Gaz, A., Giordano, R., Giroletti, S., Gobbo, B., Godang, R., Haide, I., Han, Y., Hara, K., Hayasaka, K., Hearty, C., Heidelbach, A., Higuchi, T., Himmi, A., Hoferichter, M., Howgill, D. A., Hu-Guo, C., Iijima, T., Inami, K., Irmler, C., Ishikawa, A., Itoh, R., Iyer, D., Jacobs, W. W., Jaffe, D. E., Jin, Y., Junginger, T., Kandra, J., Kojima, K., Koga, T., Korobov, A. A., Korpar, S., Križan, P., Krüger, H., Kuhr, T., Kumar, A., Kumar, R ., Kuzmin, A., Kwon, Y. -J., Lacaprara, S., Lacasta, C., Lai, Y. -T., Lalwani, K., Lam, T., Lanceri, L., Lee, M. J., Leonidopoulos, C., Levit, D., Lewis, P. M., Libby, J. F., Liu, Q. Y., Liu, Z. Y., Liventsev, D., Longo, S., Mancinelli, G., Manghisoni, M., Manoni, E., Marinas, C., Martellini, C., Martens, A., Massa, M., Massaccesi, L., Mawas, F., Mazorra, J., Merola, M., Miller, C., Minuti, M., Mizuk, R., Modak, A., Moggi, A., Mohanty, G. B., Moneta, S., Muller, Th., Na, I., Nakamura, K. R., Nakao, M., Natochii, A., Niebuhr, C., Nishida, S., Novosel, A., Pangaud, P., Parker, B., Passeri, A., Pedlar, T. K., Peinaud, Y., Peng, Y., Peschke, R., Pestotnik, R., Pham, T. H., Piccolo, M., Piilonen, L. E., Prell, S., Purohit, M. V., Ratti, L., Re, V., Reuter, L., Riceputi, E., Ripp-Baudot, I., Rizzo, G., Roney, J. M., Russo, A., Sandilya, S., Santelj, L., Savinov, V., Scavino, B., Schall, L., Schnell, G., Schwanda, C., Schwartz, A. J., Schwenker, B., Schwickardi, M., Seljak, A., Serrano, J., Shiu, J. -G., Shwartz, B., Simon, F., Soffer, A., Song, W. M., Starič, M., Stavroulakis, P., Stefkova, S., Stroili, R., Tanaka, S., Taniguchi, N., Teotia, V., Tessema, N., Thalmeier, R., Torassa, E., Trabelsi, K., Trantou, F. F., Traversi, G., Urquijo, P., Vahsen, S. E., Valin, I., Varner, G. S., Varvell, K. E., Vitale, L., Vobbilisetti, V., Wang, X. L., Wessel, C., Wienands, H. U., Won, E., Xu, D., Yamada, S., Yin, J. H., Yoshihara, K., Yuan, C. Z., Zani, L., Zong, Z., Zou, S.

We describe the planned near-term and potential longer-term upgrades of the Belle II detector at the SuperKEKB electron-positron collider operating at the KEK laboratory in Tsukuba, Japan. These upgrades will allow increasingly sensitive searches for

Externí odkaz: http://arxiv.org/abs/2406.19421

Unified Neural Backdoor Removal with Only Few Clean Samples through Unlearning and Relearning

Autor: Min, Nay Myat, Pham, Long H., Sun, Jun

The application of deep neural network models in various security-critical applications has raised significant security concerns, particularly the risk of backdoor attacks. Neural backdoors pose a serious security threat as they allow attackers to ma

Externí odkaz: http://arxiv.org/abs/2405.14781

In-situ-tunable spin-spin interactions in a Penning trap with in-bore optomechanics

Autor: Pham, Joseph H., Jee, Julian Y. Z., Rischka, Alexander, Biercuk, Michael J., Wolf, Robert N.

Publikováno v: Adv. Quantum Technol. 2024, 2400086

Experimental implementations of quantum simulation must balance the controllability of the quantum system under test with decoherence typically introduced through interaction with external control fields. The ratio of coherent interaction strength to

Externí odkaz: http://arxiv.org/abs/2401.17742