Posts
All the articles I've posted.
-
The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLMs
Published: at 11:24 AM
92.28 🤔本文通过模块化方法，利用大型语言模型参数在数学推理和多语言能力上的分离性，提出Layer-Swapping等策略，在低资源语言跨语言迁移中显著优于非模块化基线，尤其在数据受限场景下表现最佳。
-
1bit-Merging: Dynamic Quantized Merging for Large Language Models
Published: at 11:52 AM 92.20 🤔1bit-Merging提出了一种动态模型合并框架，通过1位量化任务向量和任务特定路由，在保持94.53%性能的同时将存储需求降至55.02%，在通用知识、数学推理和代码生成任务上优于传统和动态合并方法。
-
Gameplay Highlights Generation
Published: at 11:06 AM 92.19 🤔This paper presents a method to generate gameplay highlight reels by finetuning the X-CLIP multimodal model on an in-house FPS game dataset, achieving over 90% event detection accuracy and demonstrating transfer learning, while optimizing deployment through quantization.
-
Reward Reasoning Model
Published: at 11:08 AM 92.11 🤔本文提出奖励推理模型（RRMs），通过链式推理过程在生成奖励前自适应利用测试时计算资源，在多个奖励建模基准和实际应用中显著提升性能，尤其在复杂推理任务上表现优异。
-
TL;DR: Too Long, Do Re-weighting for Effcient LLM Reasoning Compression
Published: at 11:22 AM 92.10 🤔本文提出TLDR方法，通过动态再加权系统1和系统2推理数据，显著压缩大型语言模型的推理输出token数量（约40%），同时在多难度数学任务上基本保持准确性。