Zobrazeno 1 - 10
of 111
pro vyhledávání: '"Mansoor, Hassan"'
Large-language models and large-vision models are increasingly capable of solving compositional reasoning tasks, as measured by breakthroughs in visual-question answering benchmarks. However, state-of-the-art solutions often involve careful construct
Externí odkaz:
http://arxiv.org/abs/2405.19773
Autor:
Carbune, Victor, Mansoor, Hassan, Liu, Fangyu, Aralikatte, Rahul, Baechler, Gilles, Chen, Jindong, Sharma, Abhanshu
Vision-language models (VLMs) are achieving increasingly strong performance on multimodal tasks. However, reasoning capabilities remain limited particularly for smaller VLMs, while those of large-language models (LLMs) have seen numerous improvements
Externí odkaz:
http://arxiv.org/abs/2403.12596
Autor:
Sidahmed, Hakim, Phatale, Samrat, Hutcheson, Alex, Lin, Zhuonan, Chen, Zhang, Yu, Zac, Jin, Jarvis, Chaudhary, Simral, Komarytsia, Roman, Ahlheim, Christiane, Zhu, Yonghao, Li, Bowen, Ganesh, Saravanan, Byrne, Bill, Hoffmann, Jessica, Mansoor, Hassan, Li, Wei, Rastogi, Abhinav, Dixon, Lucas
While Reinforcement Learning from Human Feedback (RLHF) effectively aligns pretrained Large Language and Vision-Language Models (LLMs, and VLMs) with human preferences, its computational cost and complexity hamper its wider adoption. To alleviate som
Externí odkaz:
http://arxiv.org/abs/2403.10704
Autor:
Baechler, Gilles, Sunkara, Srinivas, Wang, Maria, Zubach, Fedir, Mansoor, Hassan, Etter, Vincent, Cărbune, Victor, Lin, Jason, Chen, Jindong, Sharma, Abhanshu
Screen user interfaces (UIs) and infographics, sharing similar visual language and design principles, play important roles in human communication and human-machine interaction. We introduce ScreenAI, a vision-language model that specializes in UI and
Externí odkaz:
http://arxiv.org/abs/2402.04615
While self-correction has shown promise in improving LLM outputs in terms of style and quality (e.g. Chen et al., 2023b; Madaan et al., 2023), recent attempts to self-correct logical or reasoning errors often cause correct answers to become incorrect
Externí odkaz:
http://arxiv.org/abs/2311.08516
Autor:
Gooding, Sian, Mansoor, Hassan
Reinforcement Learning from Human Feedback (RLHF) can be used to capture complex and nuanced properties of text generation quality. As a result, the task of text summarization has been identified as a good candidate for this process. In this paper, w
Externí odkaz:
http://arxiv.org/abs/2311.04919
Autor:
Lee, Harrison, Phatale, Samrat, Mansoor, Hassan, Mesnard, Thomas, Ferret, Johan, Lu, Kellie, Bishop, Colton, Hall, Ethan, Carbune, Victor, Rastogi, Abhinav, Prakash, Sushant
Publikováno v:
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:26874-26901, 2024
Reinforcement learning from human feedback (RLHF) has proven effective in aligning large language models (LLMs) with human preferences, but gathering high-quality preference labels is expensive. RL from AI Feedback (RLAIF), introduced in Bai et al.,
Externí odkaz:
http://arxiv.org/abs/2309.00267
Autor:
Iftekhar Ahmad, Muhammad Jalil Malik, Mansoor Hassan, Jehanzeb Ur Rehman, Shahid Abbas, Syed Mukarram Hussain
Publikováno v:
Pakistan Armed Forces Medical Journal, Vol 74, Iss 4 (2024)
Objective: To compare the incidence of port site infection with and without use of retrieval bag for extraction of gall bladder in elective laparoscopic cholecystectomy. Study Design: Quasi experimental study. Place and Duration of Study: Combin
Externí odkaz:
https://doaj.org/article/e5b367cf1aec494fb479838770a5bd26
Autor:
Mansoor, Hassan1 (AUTHOR), Lee, Isabelle Xin Yu2 (AUTHOR), Lin, Molly Tzu-Yu2 (AUTHOR), Ang, Heng Pei2 (AUTHOR), Xue, Yao Cong2 (AUTHOR), Krishaa, L.2 (AUTHOR), Patil, Moushmi2 (AUTHOR), Koh, Siew-Kwan3 (AUTHOR), Tan, Hong Chang4 (AUTHOR), Zhou, Lei5,6 (AUTHOR), Liu, Yu-Chi2,7,8,9,10 (AUTHOR) liuchiy@gmail.com
Publikováno v:
Scientific Reports. 6/11/2024, Vol. 14 Issue 1, p1-14. 14p.
Autor:
Farzana Siddique, Ashiq Hussain, Amer Ali Mahdi, Mansoor Hassan, Saima Noreen, Tahira Siddique, Syeda Ayesha Batool, Shazia Yaqub, Faiza Iftikhar Gorsi, Samina Kauser, Haya Fatima, Sameh A. Korma
Publikováno v:
Journal of Food Quality, Vol 2024 (2024)
Foods that are widely consumed and accepted have had a lot of ingredients added to them which may enhance consumer’s health. Whey is a byproduct in the manufacture of cheese or curd and is widely used in different beverage formulations, due to its
Externí odkaz:
https://doaj.org/article/64517be6ef5c47babdb46be92ed809a5