Evaluating LLM performance across long texts in five languages.
Ameeta Agrawal, Andy Dang, Sina Bagheri Nezhad
― 6 min read
Cutting edge science explained simply
Evaluating LLM performance across long texts in five languages.
Ameeta Agrawal, Andy Dang, Sina Bagheri Nezhad
― 6 min read
BEATS enhances LLMs' ability to solve math problems clearly and accurately.
Linzhuang Sun, Hao Liang, Jingxuan Wei
― 5 min read
Analyzing how gender-fair language influences language models in text classification tasks.
Andreas Waldis, Joel Birrer, Anne Lauscher
― 6 min read
PEDRO improves efficiency in fine-tuning large language models with adjustable vectors.
Tianfang Xie, Tianjing Li, Wei Zhu
― 5 min read
A new dataset to improve language models focused on business-related text.
Siyan Wang, Bradford Levy
― 5 min read
A method enhancing language model alignment with human preferences.
Zhixuan Liu, Zhanhui Zhou, Yuanfu Wang
― 5 min read
A new method improves language models' understanding of human preferences.
Jian Li, Haojing Huang, Yujia Zhang
― 4 min read
A look into assessing the trustworthiness of AI explanations through adversarial sensitivity.
Supriya Manna, Niladri Sett
― 7 min read
New method enhances AI reasoning by focusing on hierarchical semantics.
Qin Wang, Jianzhou Feng, Yiming Xu
― 5 min read
A new method enables robots to learn tasks via human interactions.
Rimvydas Rubavicius, Peter David Fagan, Alex Lascarides
― 5 min read
A new method improves search rankings using fewer examples.
Nilanjan Sinhababu, Andrew Parry, Debasis Ganguly
― 5 min read
Enhancing translation quality by aligning with human needs.
Kaden Uhlig, Joern Wuebker, Raphael Reinauer
― 6 min read
CRAFT combines RAFT and LoRA for efficient question-answering in resource-limited settings.
Isaac Chung, Phat Vo, Arman C. Kizilkale
― 6 min read
New method enhances doctor-patient communication using text and images.
Zhangpu Li, Changhong Zou, Suxue Ma
― 6 min read
Utilizing multiple annotator perspectives can improve text classification models.
Jin Xu, Mariët Theune, Daniel Braun
― 5 min read
MIPO optimizes language models by adjusting reference model influence based on data alignment.
Cheolhun Jang
― 5 min read
A new method improves logical reasoning capabilities in language models using propositional logic.
Tongxuan Liu, Wenjiang Xu, Weizhe Huang
― 5 min read
A method to improve data management in training large language models.
Hao Liang, Keshi Zhao, Yajie Yang
― 6 min read
Learn how merging AI models can improve processing speed and reduce complexity.
Stefan Hackmann
― 5 min read
Audio Moment Retrieval enables pinpointing specific moments in long recordings.
Hokuto Munakata, Taichi Nishimura, Shota Nakada
― 5 min read
Examining the role of grammar books in translating low-resource languages.
Seth Aycock, David Stap, Di Wu
― 6 min read
This article presents solutions to protect user inputs in cloud-based language services.
In Gim, Caihua Li, Lin Zhong
― 6 min read
A new method improves commonsense reasoning in AI, especially in low-data situations.
Yu Fu, Jie He, Yifan Yang
― 5 min read
AI is evolving to engage in more natural conversations.
Bandhav Veluri, Benjamin N Peloquin, Bokai Yu
― 5 min read
New methods in model training enhance reasoning abilities and efficiency.
Nikunj Saunshi, Stefani Karp, Shankar Krishnan
― 5 min read
This study automates the creation of fuzzy cognitive maps from complex social texts.
Maryam Berijanian, Spencer Dork, Kuldeep Singh
― 6 min read
Exploring transcription's impact on nursing documentation in Switzerland.
Jeremy Kramer, Tetiana Kravchenko, Beatrice Kaufmann
― 6 min read
A framework to automate and improve scientific literature summaries for researchers.
Hamed Babaei Giglou, Jennifer D'Souza, Sören Auer
― 4 min read
Examining how honesty impacts the reliability of language models.
Siheng Li, Cheng Yang, Taiqiang Wu
― 7 min read
This study uses Visual Question Answering for assessing charts created by AI models.
James Ford, Xingmeng Zhao, Dan Schumacher
― 7 min read
A project developing speech and text datasets for languages with limited resources.
Nikola Ljubešić, Peter Rupnik, Danijel Koržinek
― 5 min read
Research reveals weaknesses in online toxicity detection using ASCII art techniques.
Sergey Berezin, Reza Farahbakhsh, Noel Crespi
― 6 min read
AI is transforming scientific research with innovative technologies and applications.
Weichen Dai, Yezeng Chen, Zijie Dai
― 7 min read
Introducing an innovative method for labeling rhetorical roles in legal texts.
T. Y. S. S. Santosh, Apolline Isaia, Shiyu Hong
― 6 min read
This article explores how past cases can guide future legal decisions.
T. Y. S. S. Santosh, Mohamed Hesham Elganayni, Stanisław Sójka
― 6 min read
A survey on tasks designed for evaluating interactive agents and their capabilities.
Mareike Hartmann, Alexander Koller
― 10 min read
Exploring the performance gap of general models in finance tasks.
Yixuan Tang, Yi Yang
― 6 min read
A new method enhances model tuning efficiency using signal-to-noise ratio metrics.
Haolin Chen, Philip N. Garner
― 6 min read
New advancements in language databases enhance comparative studies across global languages.
Aditya Khan, Mason Shipton, David Anugraha
― 6 min read
A new framework enhances video-language dataset quality through iterative refinement.
Xiao Wang, Jianlong Wu, Zijia Lin
― 5 min read