Zobrazeno 1 - 10
of 1 157
pro vyhledávání: '"A Fadaee"'
Autor:
Romanou, Angelika, Foroutan, Negar, Sotnikova, Anna, Chen, Zeming, Nelaturu, Sree Harsha, Singh, Shivalika, Maheshwary, Rishabh, Altomare, Micol, Haggag, Mohamed A., A, Snegha, Amayuelas, Alfonso, Amirudin, Azril Hafizi, Aryabumi, Viraat, Boiko, Danylo, Chang, Michael, Chim, Jenny, Cohen, Gal, Dalmia, Aditya Kumar, Diress, Abraham, Duwal, Sharad, Dzenhaliou, Daniil, Florez, Daniel Fernando Erazo, Farestam, Fabian, Imperial, Joseph Marvin, Islam, Shayekh Bin, Isotalo, Perttu, Jabbarishiviari, Maral, Karlsson, Börje F., Khalilov, Eldar, Klamm, Christopher, Koto, Fajri, Krzemiński, Dominik, de Melo, Gabriel Adriano, Montariol, Syrielle, Nan, Yiyang, Niklaus, Joel, Novikova, Jekaterina, Ceron, Johan Samir Obando, Paul, Debjit, Ploeger, Esther, Purbey, Jebish, Rajwal, Swati, Ravi, Selvan Sunitha, Rydell, Sara, Santhosh, Roshan, Sharma, Drishti, Skenduli, Marjana Prifti, Moakhar, Arshia Soltani, Moakhar, Bardia Soltani, Tamir, Ran, Tarun, Ayush Kumar, Wasi, Azmine Toushik, Weerasinghe, Thenuka Ovin, Yilmaz, Serhan, Zhang, Mike, Schlag, Imanol, Fadaee, Marzieh, Hooker, Sara, Bosselut, Antoine
The performance differential of large language models (LLM) between languages hinders their effective deployment in many regions, inhibiting the potential economic and societal value of generative AI tools in many communities. However, the developmen
Externí odkaz:
http://arxiv.org/abs/2411.19799
Autor:
Gureja, Srishti, Miranda, Lester James V., Islam, Shayekh Bin, Maheshwary, Rishabh, Sharma, Drishti, Winata, Gusti, Lambert, Nathan, Ruder, Sebastian, Hooker, Sara, Fadaee, Marzieh
Reward models (RMs) have driven the state-of-the-art performance of LLMs today by enabling the integration of human feedback into the language modeling process. However, RMs are primarily trained and evaluated in English, and their capabilities in mu
Externí odkaz:
http://arxiv.org/abs/2410.15522
Autor:
Aakanksha, Ahmadian, Arash, Goldfarb-Tarrant, Seraphina, Ermis, Beyza, Fadaee, Marzieh, Hooker, Sara
Large Language Models (LLMs) have been adopted and deployed worldwide for a broad variety of applications. However, ensuring their safe use remains a significant challenge. Preference training and safety measures often overfit to harms prevalent in W
Externí odkaz:
http://arxiv.org/abs/2410.10801
Finetuning large language models on instruction data is crucial for enhancing pre-trained knowledge and improving instruction-following capabilities. As instruction datasets proliferate, selecting optimal data for effective training becomes increasin
Externí odkaz:
http://arxiv.org/abs/2409.11378
Autor:
Aryabumi, Viraat, Su, Yixuan, Ma, Raymond, Morisot, Adrien, Zhang, Ivan, Locatelli, Acyr, Fadaee, Marzieh, Üstün, Ahmet, Hooker, Sara
Including code in the pre-training data mixture, even for models not specifically designed for code, has become a common practice in LLMs pre-training. While there has been anecdotal consensus among practitioners that code data plays a vital role in
Externí odkaz:
http://arxiv.org/abs/2408.10914
The widespread adoption of synthetic data raises new questions about how models generating the data can influence other large language models (LLMs) via distilled data. To start, our work exhaustively characterizes the impact of passive inheritance o
Externí odkaz:
http://arxiv.org/abs/2407.01490
Autor:
Aakanksha, Ahmadian, Arash, Ermis, Beyza, Goldfarb-Tarrant, Seraphina, Kreutzer, Julia, Fadaee, Marzieh, Hooker, Sara
A key concern with the concept of "alignment" is the implicit question of "alignment to what?". AI systems are increasingly used across the world, yet safety alignment is often focused on homogeneous monolingual settings. Additionally, preference tra
Externí odkaz:
http://arxiv.org/abs/2406.18682
Autor:
Aryabumi, Viraat, Dang, John, Talupuru, Dwarak, Dash, Saurabh, Cairuz, David, Lin, Hangyu, Venkitesh, Bharat, Smith, Madeline, Campos, Jon Ander, Tan, Yi Chern, Marchisio, Kelly, Bartolo, Max, Ruder, Sebastian, Locatelli, Acyr, Kreutzer, Julia, Frosst, Nick, Gomez, Aidan, Blunsom, Phil, Fadaee, Marzieh, Üstün, Ahmet, Hooker, Sara
This technical report introduces Aya 23, a family of multilingual language models. Aya 23 builds on the recent release of the Aya model (\"Ust\"un et al., 2024), focusing on pairing a highly performant pre-trained model with the recently released Aya
Externí odkaz:
http://arxiv.org/abs/2405.15032
Autor:
Ahmadian, Arash, Cremer, Chris, Gallé, Matthias, Fadaee, Marzieh, Kreutzer, Julia, Pietquin, Olivier, Üstün, Ahmet, Hooker, Sara
AI alignment in the shape of Reinforcement Learning from Human Feedback (RLHF) is increasingly treated as a crucial ingredient for high performance large language models. Proximal Policy Optimization (PPO) has been positioned by recent literature as
Externí odkaz:
http://arxiv.org/abs/2402.14740
Autor:
Üstün, Ahmet, Aryabumi, Viraat, Yong, Zheng-Xin, Ko, Wei-Yin, D'souza, Daniel, Onilude, Gbemileke, Bhandari, Neel, Singh, Shivalika, Ooi, Hui-Lee, Kayid, Amr, Vargus, Freddie, Blunsom, Phil, Longpre, Shayne, Muennighoff, Niklas, Fadaee, Marzieh, Kreutzer, Julia, Hooker, Sara
Recent breakthroughs in large language models (LLMs) have centered around a handful of data-rich languages. What does it take to broaden access to breakthroughs beyond first-class citizen languages? Our work introduces Aya, a massively multilingual g
Externí odkaz:
http://arxiv.org/abs/2402.07827