Breaking Babel: A Self-Evolving Multi-Agent System for Long-Form Subtitle Translation

类型：arxiv
标识：2609.38660
链接：https://arxiv.org/abs/2609.38660
主分类：agent
形态：method
TLDR：Long-form subtitle translation requires reasoning over discourse and cultural context spanning episodes or entire series, while maintaining consistent terminology and style. Existing single-LLM methods are largely sentence-level, and multi-agent systems often use static workflows that do not adapt to scene complexity or production context. We propose SMART, a Self-evolving Multi-Agent system for long-foRm subtitle Translation. During test-time training, SMART builds persistent series-level memory and translates a subset of sentences through a dynamic router and Mixture-of-Agents layer with too
待LLM分类：否
来源文件：
/inbox/tom/_candidates/2026-10-01-agent-rag-longcontext-candidates.json

2026-09-01

笔记 flyP 精读与批判 v2 · VLD-RAG · Agentic Vision-Language Document RAG for Long Multi-Page · 重写覆盖（2026-09-25 → 2026-09-26）

2026-09-25

笔记 Self-Challenging Language Model Agents (SCA) + Code-as-Task · flyP 单篇精读（2026-08-18 22:50 · 晚棒）

2026-08-18

笔记 flyP 精读与批判 · 2026-07-29 22:50（v2 覆盖）

2026-07-29

仓库 afif-pajar/indo-caption-pipeline-to-rag-for-learning-source

2026-08-11

仓库 thuyhuongctu/M-AIDA

2026-09-29

arXiv

2609.38660

TLDR

Long-form subtitle translation requires reasoning over discourse and cultural context spanning episodes or entire series, while maintaining consistent terminology and style. Existing single-LLM methods are largely sentence-level, and multi-agent systems often use static workflows that do not adapt to scene complexity or production context. We propose SMART, a Self-evolving Multi-Agent system for long-foRm subtitle Translation. During test-time training, SMART builds persistent series-level memory and translates a subset of sentences through a dynamic router and Mixture-of-Agents layer with too

Breaking Babel: A Self-Evolving Multi-Agent System for Long-Form Subtitle Translation

相关 · 按语义匹配