Results 271 to 280 of about 152,555 (310)

Critical Tokens Matter: Token-Level Contrastive Estimation Enhances LLM's Reasoning Capability

International Conference on Machine Learning
Mathematical reasoning tasks pose significant challenges for large language models (LLMs) because they require precise logical deduction and sequence analysis.
Zicheng Lin, Tian Liang, Jiahao Xu, Xing Wang, Ruilin Luo, Chufan Shi, Siheng Li, Yujiu Yang, Zhaopeng Tu   +8 more
semanticscholar   +1 more source

Keypoint Action Tokens Enable In-Context Imitation Learning in Robotics

Robotics: Science and Systems Conference
We show that off-the-shelf text-based Transformers, with no additional training, can perform few-shot in-context visual imitation learning, mapping visual observations to action sequences that emulate the demonstrator's behaviour.
Norman Di Palo, Edward Johns
semanticscholar   +1 more source

Perception Tokens Enhance Visual Reasoning in Multimodal Language Models

Computer Vision and Pattern Recognition
Multimodal language models (MLMs) still face challenges in fundamental visual perception tasks where specialized models excel. Tasks requiring reasoning about 3D structures benefit from depth estimation, and reasoning about 2D object instances benefits ...
Mahtab Bigverdi, Zelun Luo, Cheng-Yu Hsieh, Ethan Shen, Dongping Chen, Linda G. Shapiro, Ranjay Krishna   +6 more
semanticscholar   +1 more source

Not All Tokens Are What You Need for Pretraining

Neural Information Processing Systems
Previous language model pre-training methods have uniformly applied a next-token prediction loss to all training tokens. Challenging this norm, we posit that “Not all tokens in a corpus are equally important for language model training” .
Zheng-Wen Lin, Zhibin Gou, Yeyun Gong, Xiao Liu, Yelong Shen, Ruochen Xu, Chen Lin, Yujiu Yang, Jian Jiao, Nan Duan, Weizhu Chen   +10 more
semanticscholar   +1 more source

DeepStack: Deeply Stacking Visual Tokens is Surprisingly Simple and Effective for LMMs

Neural Information Processing Systems
Most large multimodal models (LMMs) are implemented by feeding visual tokens as a sequence into the first layer of a large language model (LLM). The resulting architecture is simple but significantly increases computation and memory costs, as it has to ...
Lingchen Meng, Jianwei Yang, Rui Tian, Xiyang Dai, Zuxuan Wu, Jianfeng Gao, Yu-Gang Jiang   +6 more
semanticscholar   +1 more source

Tokenomics: When Tokens Beat Equity

Management Science, 2023
Katya Malinová, Andreas Park
exaly  

Prospecting non-fungible tokens in the digital economy: Stakeholders and ecosystem, risk and opportunity

Business Horizons, 2022
Hadi Ghaderi
exaly  

Interconnectedness between healthcare tokens and healthcare stocks: Evidence from a quantile VAR approach

International Review of Economics and Finance, 2023
Imran Yousaf, John W Goodell
exaly  

Extreme connectedness between renewable energy tokens and fossil fuel markets

Energy Economics, 2022
Muhammad Umar, Ramzi Nekhili, Imran Yousaf   +2 more
exaly  

Connectedness among fan tokens and stocks of football clubs

Research in International Business and Finance, 2022
Ata Assaf, Oguz Ersan, Ender Demir
exaly