Kimi k1. 5: Scaling reinforcement learning with llms
Language model pretraining with next token prediction has proved effective for scaling compute
but is limited to the amount of available training data. Scaling reinforcement learning (RL…
but is limited to the amount of available training data. Scaling reinforcement learning (RL…
Kimi-vl technical report
We present Kimi-VL, an efficient open-source Mixture-of-Experts (MoE) vision-language
model (VLM) that offers advanced multimodal reasoning, long-context understanding, and …
model (VLM) that offers advanced multimodal reasoning, long-context understanding, and …
Incorporating Dense Knowledge Alignment into Unified Multimodal Representation Models
Y Cui, X Zu, W Zhang, Z Zhao… - Proceedings of the …, 2025 - openaccess.thecvf.com
Leveraging Large Language Models (LLMs) for text representation has achieved significant
success, but the exploration of using Multimodal LLMs (MLLMs) for multimodal …
success, but the exploration of using Multimodal LLMs (MLLMs) for multimodal …
Onestop qamaker: extract question-answer pairs from text in a one-stop approach
S Cui, X Bao, X Zu, Y Guo, Z Zhao, J Zhang… - arXiv preprint arXiv …, 2021 - arxiv.org
Large-scale question-answer (QA) pairs are critical for advancing research areas like machine
reading comprehension and question answering. To construct QA pairs from documents …
reading comprehension and question answering. To construct QA pairs from documents …
A two-stage conversational query rewriting model with multi-task learning
Conversational context understanding aims to recognize the real intention of user from the
conversation history, which is critical for building the dialogue system. However, the multi-turn …
conversation history, which is critical for building the dialogue system. However, the multi-turn …
VK-G2T: Vision and Context Knowledge Enhanced Gloss2text
Existing sign language translation methods follow a two-stage pipeline: first converting the
sign language video to a gloss sequence (ie, Sign2Gloss) and then translating the generated …
sign language video to a gloss sequence (ie, Sign2Gloss) and then translating the generated …
MLR: A Two-stage Conversational Query Rewriting Model with Multi-task Learning
Conversational context understanding aims to recognize the real intention of user from the
conversation history, which is critical for building the dialogue system. However, the multi-turn …
conversation history, which is critical for building the dialogue system. However, the multi-turn …
A method of time synchronization of uplink signal on leo satellite
L Ning-rui, C Jian, G Xin-xing… - … Conference on Wireless …, 2015 - ieeexplore.ieee.org
… We can get four pseudo range equations about four unknown quantities (xu, yu, zu, ∆t) from
at least four measurement. The user coordinates (xu, yu, zu) and clock offset (∆t) could be …
at least four measurement. The user coordinates (xu, yu, zu) and clock offset (∆t) could be …
Treatment of sagittal fracture of the mandibular condyle using resorbable-screw osteosynthesis
X Xin, Y Zhao, G Cheng, D Diarra, ZB Li, Z Li - Journal of Oral and …, 2022 - Elsevier
Purpose Screw osteosynthesis is advocated for the treatment of sagittal fracture of
mandibular condyle (SFMC). This study aimed to explore the applicability of resorbable-screw …
mandibular condyle (SFMC). This study aimed to explore the applicability of resorbable-screw …
Unusual threshold anomaly in the 6li+ 208pb system
…, J Hui-Ming, W Zhen-Dong, X Xin-Xing… - Chinese Physics …, 2006 - iopscience.iop.org
The angular distributions of elastic scattering for the 6 Li+ 208 Pb system have been measured
at several energies around the Coulomb barrier. The parameters of optical potential are …
at several energies around the Coulomb barrier. The parameters of optical potential are …