What should a neuron aim for? Designing local objective functions based on information theory

Autor: Schneider, Andreas C., Neuhaus, Valentin, Ehrlich, David A., Makkeh, Abdullah, Ecker, Alexander S., Priesemann, Viola, Wibral, Michael

In modern deep neural networks, the learning dynamics of the individual neurons is often obscure, as the networks are trained via global optimization. Conversely, biological systems build on self-organized, local learning, achieving robustness and ef

Externí odkaz: http://arxiv.org/abs/2412.02482

Measurement of the double-differential cross section of muon-neutrino charged-current interactions with low hadronic energy in the NOvA Near Detector

Autor: Acero, M. A., Acharya, B., Adamson, P., Aliaga, L., Anfimov, N., Antoshkin, A., Arrieta-Diaz, E., Asquith, L., Aurisano, A., Back, A., Balashov, N., Baldi, P., Bambah, B. A., Bannister, E., Barros, A., Bashar, S., Bat, A., Bays, K., Bernstein, R., Bezerra, T. J. C., Bhatnagar, V., Bhattarai, D., Bhuyan, B., Bian, J., Booth, A. C., Bowles, R., Brahma, B., Bromberg, C., Buchanan, N., Butkevich, A., Calvez, S., Carroll, T. J., Catano-Mur, E., Cesar, J. P., Chatla, A., Chirco, R., Choudhary, B. C., Christensen, A., Cicala, M. F., Coan, T. E., Cooleybeck, A., Cortes-Parra, C., Coveyou, D., Cremonesi, L., Davies, G. S., Derwent, P. F., Ding, P., Djurcic, Z., Dobbs, K., Dolce, M., Doyle, D., Tonguino, D. Dueñas, Dukes, E. C., Dye, A., Ehrlich, R., Ewart, E., Filip, P., Frank, M. J., Gallagher, H. R., Gao, F., Giri, A., Gomes, R. A., Goodman, M. C., Groh, M., Group, R., Habig, A., Hakl, F., Hartnell, J., Hatcher, R., He, M., Heller, K., Hewes, V, Himmel, A., Horoho, T., Ivaneev, Y., Ivanova, A., Jargowsky, B., Jarosz, J., Johnson, C., Judah, M., Kakorin, I., Kaplan, D. M., Kalitkina, A., Kirezli-Ozdemir, B., Kleykamp, J., Klimov, O., Koerner, L. W., Kolupaeva, L., Kralik, R., Kumar, A., Kus, V., Lackey, T., Lang, K., Lesmeister, J., Lister, A., Liu, J., Lock, J. A., Lokajicek, M., MacMahon, M., Magill, S., Mann, W. A., Manoharan, M. T., Plata, M. Manrique, Marshak, M. L., Martinez-Casales, M., Matveev, V., Mehta, B., Messier, M. D., Meyer, H., Miao, T., Miller, W. H., Mishra, S., Mishra, S. R., Mislivec, A., Mohanta, R., Moren, A., Morozova, A., Mu, W., Mualem, L., Muether, M., Mulder, K., Myers, D., Naples, D., Nath, A., Nelleri, S., Nelson, J. K., Nichol, R., Niner, E., Norman, A., Norrick, A., Nosek, T., Oh, H., Olshevskiy, A., Olson, T., Ozkaynak, M., Pal, A., Paley, J., Panda, L., Patterson, R. B., Pawloski, G., Petti, R., Porter, J. C. C., Prais, L. R., Rabelhofer, M., Rafique, A., Raj, V., Rajaoalisoa, M., Ramson, B., Rebel, B., Roy, P., Samoylov, O., Sanchez, M. C., Falero, S. Sanchez, Shanahan, P., Sharma, P., Sheshukov, A., Shivam, Shmakov, A., Shorrock, W., Shukla, S., Singha, D. K., Singh, I., Singh, P., Singh, V., Smith, E., Smolik, J., Snopok, P., Solomey, N., Sousa, A., Soustruznik, K., Strait, M., Suter, L., Sutton, A., Sutton, K., Swain, S., Sweeney, C., Sztuc, A., Talukdar, N., Oregui, B. Tapia, Tas, P., Thakore, T., Thomas, J., Tiras, E., Titus, M., Torun, Y., Tran, D., Trokan-Tenorio, J., Urheim, J., Vahle, P., Vallari, Z., Villamil, J. D., Vockerodt, K. J., Wallbank, M., Weber, C., Wetstein, M., Whittington, D., Wickremasinghe, D. A., Wieber, T., Wolcott, J., Wrobel, M., Wu, S., Wu, W., Xiao, Y., Yaeggy, B., Yahaya, A., Yankelevich, A., Yonehara, K., Yu, Y., Zadorozhnyy, S., Zalesak, J., Zwaska, R.

The NOvA collaboration reports cross-section measurements for $\nu_{\mu}$ charged-current interactions with low hadronic energy (maximum kinetic energy of 250 MeV for protons and 175 MeV for pions) in the NOvA Near Detector. The results are presented

Externí odkaz: http://arxiv.org/abs/2410.10222

Latent-INR: A Flexible Framework for Implicit Representations of Videos with Discriminative Semantics

Autor: Maiya, Shishira R, Gupta, Anubhav, Gwilliam, Matthew, Ehrlich, Max, Shrivastava, Abhinav

Implicit Neural Networks (INRs) have emerged as powerful representations to encode all forms of data, including images, videos, audios, and scenes. With video, many INRs for video have been proposed for the compression task, and recent methods featur

Externí odkaz: http://arxiv.org/abs/2408.02672

Wolf: Captioning Everything with a World Summarization Framework

Autor: Li, Boyi, Zhu, Ligeng, Tian, Ran, Tan, Shuhan, Chen, Yuxiao, Lu, Yao, Cui, Yin, Veer, Sushant, Ehrlich, Max, Philion, Jonah, Weng, Xinshuo, Xue, Fuzhao, Tao, Andrew, Liu, Ming-Yu, Fidler, Sanja, Ivanovic, Boris, Darrell, Trevor, Malik, Jitendra, Han, Song, Pavone, Marco

We propose Wolf, a WOrLd summarization Framework for accurate video captioning. Wolf is an automated captioning framework that adopts a mixture-of-experts approach, leveraging complementary strengths of Vision Language Models (VLMs). By utilizing bot

Externí odkaz: http://arxiv.org/abs/2407.18908

Hydragen: High-Throughput LLM Inference with Shared Prefixes

Autor: Juravsky, Jordan, Brown, Bradley, Ehrlich, Ryan, Fu, Daniel Y., Ré, Christopher, Mirhoseini, Azalia

Transformer-based large language models (LLMs) are now deployed to hundreds of millions of users. LLM inference is commonly performed on batches of sequences that share a prefix, such as few-shot examples or a chatbot system prompt. Decoding in this

Externí odkaz: http://arxiv.org/abs/2402.05099