Zobrazeno 1 - 10
of 7 893
pro vyhledávání: '"Dingli, A."'
Autor:
Abdin, Marah, Aneja, Jyoti, Behl, Harkirat, Bubeck, Sébastien, Eldan, Ronen, Gunasekar, Suriya, Harrison, Michael, Hewett, Russell J., Javaheripi, Mojan, Kauffmann, Piero, Lee, James R., Lee, Yin Tat, Li, Yuanzhi, Liu, Weishung, Mendes, Caio C. T., Nguyen, Anh, Price, Eric, de Rosa, Gustavo, Saarikivi, Olli, Salim, Adil, Shah, Shital, Wang, Xin, Ward, Rachel, Wu, Yue, Yu, Dingli, Zhang, Cyril, Zhang, Yi
We present phi-4, a 14-billion parameter language model developed with a training recipe that is centrally focused on data quality. Unlike most language models, where pre-training is based primarily on organic data sources such as web content or code
Externí odkaz:
http://arxiv.org/abs/2412.08905
Multiple-input multiple-output (MIMO) is pivotal for wireless systems, yet its high-dimensional, stochastic channel poses significant challenges for accurate estimation, highlighting the critical need for robust estimation techniques. In this paper,
Externí odkaz:
http://arxiv.org/abs/2410.23752
As large language models (LLMs) become increasingly advanced, their ability to exhibit compositional generalization -- the capacity to combine learned skills in novel ways not encountered during training -- has garnered significant attention. This ty
Externí odkaz:
http://arxiv.org/abs/2409.19808
Compositionality is a critical capability in Text-to-Image (T2I) models, as it reflects their ability to understand and combine multiple concepts from text descriptions. Existing evaluations of compositional capability rely heavily on human-designed
Externí odkaz:
http://arxiv.org/abs/2408.14339
Autor:
Shah, Vedant, Yu, Dingli, Lyu, Kaifeng, Park, Simon, Yu, Jiatong, He, Yinghui, Ke, Nan Rosemary, Mozer, Michael, Bengio, Yoshua, Arora, Sanjeev, Goyal, Anirudh
Current LLM training positions mathematical reasoning as a core capability. With publicly available sources fully tapped, there is unmet demand for diverse and challenging math questions. Relying solely on human experts is both time-consuming and cos
Externí odkaz:
http://arxiv.org/abs/2407.21009
Although Reinforcement Learning (RL) algorithms acquire sequential behavioral patterns through interactions with the environment, their effectiveness in noisy and high-dimensional scenarios typically relies on specific structural priors. In this pape
Externí odkaz:
http://arxiv.org/abs/2404.09760
Public LLMs such as the Llama 2-Chat have driven huge activity in LLM research. These models underwent alignment training and were considered safe. Recently Qi et al. (2023) reported that even benign fine-tuning (e.g., on seemingly safe datasets) can
Externí odkaz:
http://arxiv.org/abs/2402.18540
We investigate coherency properties of certain completed integral group rings, precisely for compact $p$-adic Lie groups.
Comment: 16 pages. Submitted
Comment: 16 pages. Submitted
Externí odkaz:
http://arxiv.org/abs/2401.05506
We prove a general structure theorem for finitely presented torsion modules over a class of commutative rings that need not be Noetherian. As a first application, we then use this result to study the Weil- \'etale cohomology groups of $\mathbb{G}_m$
Externí odkaz:
http://arxiv.org/abs/2401.02946
Autor:
Shen, Bin, Xia, Dingli
In this manuscript, we study the positive solutions of the Finslerian Fisher-KPP equation $$ u_t=\Delta^{\nabla u} u+cu(1-u). $$ The Fisher-KPP equation is widely applied and connected to many mathematical branches. We establish the global gradient e
Externí odkaz:
http://arxiv.org/abs/2403.00002