Zobrazeno 1 - 10
of 27
pro vyhledávání: '"Somepalli, Gowthami"'
Autor:
Singla, Vasu, Yue, Kaiyu, Paul, Sukriti, Shirkavand, Reza, Jayawardhana, Mayuka, Ganjdanesh, Alireza, Huang, Heng, Bhatele, Abhinav, Somepalli, Gowthami, Goldstein, Tom
Training large vision-language models requires extensive, high-quality image-text pairs. Existing web-scraped datasets, however, are noisy and lack detailed image descriptions. To bridge this gap, we introduce PixelProse, a comprehensive dataset of o
Externí odkaz:
http://arxiv.org/abs/2406.10328
Autor:
Hans, Abhimanyu, Wen, Yuxin, Jain, Neel, Kirchenbauer, John, Kazemi, Hamid, Singhania, Prajwal, Singh, Siddharth, Somepalli, Gowthami, Geiping, Jonas, Bhatele, Abhinav, Goldstein, Tom
Large language models can memorize and repeat their training data, causing privacy and copyright risks. To mitigate memorization, we introduce a subtle modification to the next-token training objective that we call the goldfish loss. During training,
Externí odkaz:
http://arxiv.org/abs/2406.10209
Autor:
Rawal, Ruchit, Saifullah, Khalid, Farré, Miquel, Basri, Ronen, Jacobs, David, Somepalli, Gowthami, Goldstein, Tom
Current datasets for long-form video understanding often fall short of providing genuine long-form comprehension challenges, as many tasks derived from these datasets can be successfully tackled by analyzing just one or a few random frames from a vid
Externí odkaz:
http://arxiv.org/abs/2405.08813
Autor:
Kirchenbauer, John, Honke, Garrett, Somepalli, Gowthami, Geiping, Jonas, Ippolito, Daphne, Lee, Katherine, Goldstein, Tom, Andre, David
We develop a methodology for analyzing language model task performance at the individual example level based on training data density estimation. Experiments with paraphrasing as a controlled intervention on finetuning data demonstrate that increasin
Externí odkaz:
http://arxiv.org/abs/2405.06331
Autor:
Somepalli, Gowthami, Gupta, Anubhav, Gupta, Kamal, Palta, Shramay, Goldblum, Micah, Geiping, Jonas, Shrivastava, Abhinav, Goldstein, Tom
Generative models are now widely used by graphic designers and artists. Prior works have shown that these models remember and often replicate content from their training data during generation. Hence as their proliferation increases, it has become im
Externí odkaz:
http://arxiv.org/abs/2404.01292
Autor:
Cherepanova, Valeriia, Levin, Roman, Somepalli, Gowthami, Geiping, Jonas, Bruss, C. Bayan, Wilson, Andrew Gordon, Goldstein, Tom, Goldblum, Micah
Publikováno v:
Conference on Neural Information Processing Systems 2023
Academic tabular benchmarks often contain small sets of curated features. In contrast, data scientists typically collect as many features as possible into their datasets, and even engineer new features from existing ones. To prevent overfitting in su
Externí odkaz:
http://arxiv.org/abs/2311.05877
Autor:
Goldblum, Micah, Souri, Hossein, Ni, Renkun, Shu, Manli, Prabhu, Viraj, Somepalli, Gowthami, Chattopadhyay, Prithvijit, Ibrahim, Mark, Bardes, Adrien, Hoffman, Judy, Chellappa, Rama, Wilson, Andrew Gordon, Goldstein, Tom
Neural network based computer vision systems are typically built on a backbone, a pretrained or randomly initialized feature extractor. Several years ago, the default option was an ImageNet-trained convolutional neural network. However, the recent pa
Externí odkaz:
http://arxiv.org/abs/2310.19909
Autor:
Jain, Neel, Chiang, Ping-yeh, Wen, Yuxin, Kirchenbauer, John, Chu, Hong-Min, Somepalli, Gowthami, Bartoldson, Brian R., Kailkhura, Bhavya, Schwarzschild, Avi, Saha, Aniruddha, Goldblum, Micah, Geiping, Jonas, Goldstein, Tom
We show that language model finetuning can be improved, sometimes dramatically, with a simple augmentation. NEFTune adds noise to the embedding vectors during training. Standard finetuning of LLaMA-2-7B using Alpaca achieves 29.79% on AlpacaEval, whi
Externí odkaz:
http://arxiv.org/abs/2310.05914
Autor:
Jain, Neel, Schwarzschild, Avi, Wen, Yuxin, Somepalli, Gowthami, Kirchenbauer, John, Chiang, Ping-yeh, Goldblum, Micah, Saha, Aniruddha, Geiping, Jonas, Goldstein, Tom
As Large Language Models quickly become ubiquitous, it becomes critical to understand their security vulnerabilities. Recent work shows that text optimizers can produce jailbreaking prompts that bypass moderation and alignment. Drawing from the rich
Externí odkaz:
http://arxiv.org/abs/2309.00614
Images generated by diffusion models like Stable Diffusion are increasingly widespread. Recent works and even lawsuits have shown that these models are prone to replicating their training data, unbeknownst to the user. In this paper, we first analyze
Externí odkaz:
http://arxiv.org/abs/2305.20086