研究库 论文知识库

内容库 / 论文

arXiv:2609.25804 · Agent 智能体

The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks

The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks

类型：arxiv
标识：2609.25804
链接：https://arxiv.org/abs/2609.25804
主分类：agent
形态：benchmark
TLDR：LLM agents increasingly work on long-horizon tasks, and the decisions they make along the way, such as which hypothesis to test or which implementation to build on, determine the outcome of the whole run. Making these decisions well is becoming a key capability for both engineering and research agents. We refer to the ability to make good long-horizon decisions as the taste of an agent. While existing benchmarks measure the end-to-end success of agents on long-horizon tasks, none of them measures the taste of an agent. To address this problem, we build Taste-Bench, a benchmark of taste questio
副分类：evaluation
待LLM分类：否
来源文件：
/inbox/tom/_candidates/2026-09-23-agent-rag-longcontext-candidates.json

相关 · 按语义匹配

笔记 Tom 文献雷达 · Agent + RAG + LongContext · 2026-08-29

笔记 Tom 文献雷达 · Agent RAG LongContext · 2026-09-04 08:40

笔记 📋 Tom 文献雷达 · Agent + RAG + 长上下文 · 2026-09-01

文档 Tom 评 flyP · 2026-07-24 · Self Gradient Forcing + Anchor-Align 精读

笔记 📋 Tom 文献雷达 · Agent + RAG + Long Context · 2026-08-28 14:40 UTC

笔记 Tom 文献雷达 · AI Agent / RAG / 长上下文 · 2026-07-10

arXiv

2609.25804

TLDR

LLM agents increasingly work on long-horizon tasks, and the decisions they make along the way, such as which hypothesis to test or which implementation to build on, determine the outcome of the whole run. Making these decisions well is becoming a key capability for both engineering and research agents. We refer to the ability to make good long-horizon decisions as the taste of an agent. While existing benchmarks measure the end-to-end success of agents on long-horizon tasks, none of them measures the taste of an agent. To address this problem, we build Taste-Bench, a benchmark of taste questio

arXiv ↗ 中文全文