大模型算法工程师 | 专注于评测与 Judge Model 构建
LLM Algorithm Engineer | Focused on Evaluation & Judge Model
我热衷于探索大模型的边界，不仅关注模型如何生成，更关注如何客观、高效地评价生成质量。
I'm passionate about exploring the boundaries of LLMs, focusing not only on generation but also on objective and efficient evaluation.
| 领域 | 中文描述 | Description |
|---|---|---|
| Judge Model | 研究如何构建更公正、更符合人类偏好的裁判模型 | Building fairer, human-preference-aligned judge models |
| Evaluation | 关注模型幻觉、长文本能力与复杂逻辑推理的自动评测 | Evaluating hallucination, long-context capabilities & complex reasoning |
| Agent Architecture | 对新兴 Agent 架构与改变规则的开源项目保持高度关注 | Tracking emerging Agent architectures and game-changing open-source projects |
| 项目 | 技术栈 | 简介 |
|---|---|---|
| leelicspace | Next.js · TypeScript · Tailwind | 个人博客系统，支持 Markdown、RSS、暗色模式 / Personal blog with admin dashboard & RSS |
| 图文排版助手 Pro | React · Tailwind · html2canvas | 社交媒体卡片生成器，支持 Markdown 与长文分页 / Social media card creator with Markdown support |
| SpeakSense | FastAPI · Python · Whisper | AI 演讲分析工具，检测口头禅与语速 / AI speech analysis for filler detection & pace |
| note | - | 每周学习笔记 / Weekly learning notes |
AI & ML
Web
