Rulin Shao RulinShao

## README.html
<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1.0" />
  <title>ServingKernelEvolveBench (SKB)</title>
  <meta name="description" content="A decomposable, dense-reward benchmark for evolutionary / RL agents that optimize an end-to-end LLM serving stack." />
  <meta property="og:title" content="ServingKernelEvolveBench (SKB)" />
  <meta property="og:description" content="A decomposable, dense-reward benchmark for evolutionary / RL agents that optimize an end-to-end LLM serving stack." />
  <meta property="og:type" content="website" />

## dr_tulu_sft_curation_prompt.py
from textwrap import dedent

AGENT_SYSTEM_PROMPT_0902 = dedent(r"""
You are a research assistant that answers questions through iterative reasoning and evidence-backed search using multiple external search systems.

1. Operating Principles, Process & Guidelines

1.1 Principles

- Provide comprehensive, evidence-backed answers to scientific questions.
	<!DOCTYPE html>
	<html lang="en">
	<head>
	<meta charset="UTF-8" />
	<meta name="viewport" content="width=device-width, initial-scale=1.0" />
	<title>ServingKernelEvolveBench (SKB)</title>
	<meta name="description" content="A decomposable, dense-reward benchmark for evolutionary / RL agents that optimize an end-to-end LLM serving stack." />
	<meta property="og:title" content="ServingKernelEvolveBench (SKB)" />
	<meta property="og:description" content="A decomposable, dense-reward benchmark for evolutionary / RL agents that optimize an end-to-end LLM serving stack." />
	<meta property="og:type" content="website" />
	from textwrap import dedent

	AGENT_SYSTEM_PROMPT_0902 = dedent(r"""
	You are a research assistant that answers questions through iterative reasoning and evidence-backed search using multiple external search systems.

	1. Operating Principles, Process & Guidelines

	1.1 Principles

	- Provide comprehensive, evidence-backed answers to scientific questions.