慣性聚合 高效追讀感興趣之博客、新聞、科技資訊
閱原文 以慣性聚合開啟

推薦訂閱源

Google DeepMind News
Google DeepMind News
人人都是产品经理
人人都是产品经理
M
MIT News - Artificial intelligence
博客园 - 叶小钗
MyScale Blog
MyScale Blog
V
Visual Studio Blog
月光博客
月光博客
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
量子位
I
InfoQ
有赞技术团队
有赞技术团队
阮一峰的网络日志
阮一峰的网络日志
Jina AI
Jina AI
V
V2EX
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
Blog — PlanetScale
Blog — PlanetScale
Last Week in AI
Last Week in AI
雷峰网
雷峰网
Stack Overflow Blog
Stack Overflow Blog
博客园 - Franky

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant Common SOC 2 Failures (Real World) Stop Vibe-Checking Your AI App: A Practical Guide to Evals How to Use SonarQube and SonarScanner Locally to Level Up Your Code Quality Your Next To-Do App Is Dead — I Replaced Mine with an OpenClaw AI Sign a Nostr event in 60 lines of Python using coincurve — no nostr-sdk, no nbxplorer, no rust toolchain ITGC Audit Explained Like You’re in Big 4 Patch Tuesday abril 2026: Microsoft parcha 163 vulnerabilidades y un zero-day en SharePoint Stop scraping everything: a better way to track competitor price changes Listing on MCPize + the Official MCP Registry while routing payments OUTSIDE the marketplace — how I kept 100% of my x402 revenue Building an AI-Powered Risk Intelligence System Using Serverless Architecture Why We Ripped Function Overloading Out of Our AI Toolchain Testing AI-Generated Code: How to Actually Know If It Works SaaS Churn Is Killing Your Business. Here Is What to Do About It (Without a Support Team) The Speed of AI Is No Longer Linear - And Self-Improving Models Are Why How to Implement RBAC for MCP Tools: A Practical Guide for Engineering Teams From Standard Quote to Persuasive Proposal: AI Automation for Arborists I built a CLI that scaffolds complete multi-tenant SaaS apps Axios CVE-2025–62718: The Silent SSRF Bug That Could Be Hiding in Your Node.js App Right Now The dashboard that ended our friendship Data Pipelines Explained Simply (and How to Build Them with Python)
善进人工智能简历匹配之法,以提示迭代——自七点三七至八点三七分。
Azeez Roheem · 2026-05-25 · via DEV Community

众人所讳言者

此乃多数AI简历润色者遇弱项时之所为

原文"日赴站会"

AI润色"以React及TypeScript构建响应式网页应用,协于敏捷站会,以成高质量前端之解"

是也。有要言,有结构,闻之专业。

然亦妄也。

是人也,未尝开React之应用。惟会而已。AI观其职说,见React与TypeScript,默造其无之经验。若得面试,首问必现其伪。

此非虚言。此乃吾之简历改写器——Resume AI Tailor——未修正前所为也。吾知之,盖量之故也。

Resume AI Tailor何为

Resume AI Tailor者,乃吾于NanoCrafts所建之SaaS产品也。上传简历,粘贴职位描述,秒得定制版本。其技术栈为Next.js、gpt-4o-mini、Clerk、Neon/Drizzle及Vercel。

其本旨明矣:履历愈优,匹配分愈高,面试机愈广。然此旨,一俟人工智能妄造应聘者所无之技,便立破矣。

以重写履历者,为Wordpress开发者杜撰React之经,非助其益,实设其败于面试之堂。

吾之所为者

余用两周之时,研习进阶之提示工程术——少样本提示、角色提示,及以大语言模型为判官之评估循环——专攻简历AI定制器之重写与分析路径。

其志非在令输出更美,而在使其可量度之善,随提示之变,有数可循。

吾初建评鉴之环:复以模型呼出,量每改写之要目,依五则(行动词、关键词契合、成果、真确、简练)而评,循定式之标。继而以十份履历过此管道,立为基准。

初周之基:三十八要目,得七有三十七分。

复行之。二旬,屡易提示,遍测五履历之境。每变皆度此基线。

终得分数:8.37/10。净进:增1.0分。

是文记所变者何,何以效,及所悟非在任一提示工程之典。

少举提示:示之勿言。

原之改写提示若此:

const prompt = `You are an expert resume writer. Rewrite the following 
bullet point to be stronger, specific, and results-oriented.

Original bullet: "${bullet}"
Rewritten bullet:`;

入全屏模式 出全屏模式

此乃零样本提示。一令,无例。模型知"强"之抽象义,然不知软件工程师求高阶之职,简历标目中"强"之具体形何如。

所出之文,自信其声,然结构不谐。或为段落,或为点列,或模型增以指标,或竟无之。复有之时,竟幻生技能,全然无据。

少样本提示之法,以示所求之文,而非述之,故能正其格式之弊。

const REWRITE_SYSTEM_PROMPT = `You are a senior ATS engineer and resume 
specialist with 10 years of experience configuring applicant tracking 
systems at Fortune 500 companies.

Study the examples below carefully. Your output must match this format 
exactly — return only the rewritten bullet, no explanation.

EXAMPLE 1:

Original: "Worked on the backend of the company's main product"
Target keywords: Node.js, REST APIs, microservices
Job title: Senior Software Engineer
Candidate skills: Node.js, PostgreSQL, Docker, REST APIs

Rewritten: Engineered Node.js REST APIs for core product, improving 
system performance across microservices architecture

EXAMPLE 2:

Original: "Attended daily standups"
Target keywords: React, TypeScript, agile, frontend
Job title: Frontend Developer
Candidate skills: HTML, CSS, JavaScript, WordPress

Rewritten: Participated in agile daily standups supporting [React] 
and [TypeScript] frontend delivery

Return ONLY the rewritten bullet. No explanation. Maximum 20 words.`;

入全屏模式 出全屏模式

是三者同时发于彼示:

其职引其注意。"资深ATS工程师"示模型所当取之训——关键词之遵,格式之严,解析之易。非泛泛之文质。

其例定其式。 举凡运行,皆得单行输出,其式如所示,无段落,无编号,无赘言.

例二示诚. 候选人具HTML、CSS、JavaScript、WordPress之能,非React或TypeScript。例显模型当如何处所缺之技:宜用方括号占位,如[React],毋直言技能。

末节为要。单凭指令未解幻象之困。示以模型正确占位之实例方解之。

要旨所在:简例与书示相悖,则简例为重。 若诸例皆显应者自信其技之娴熟,则模型将循例而行,非从"须诚"之令。示汝所求。慎其所示。

角色提示:授模型以镜。

角色提示,乃于用户内容未处理前,以系统提示为模型赋予人格。模型非成其人,惟变其所本重之训练数据而已。

此别甚关。言"汝为招募之长"者,乃注目之指令,非能力之升。模型本已知招募之长所重,今惟令其重此知于万钧之上。

余试三人于次日,同以一简牍相较。

ATS工程师

`You are a senior ATS engineer with 10 years of experience configuring 
applicant tracking systems at Fortune 500 companies like Greenhouse, 
Workday, and Lever.

When analysing a resume, evaluate it exactly as an ATS would before 
a human ever sees it. Your priorities are:

1. Keyword match — does the resume contain exact terms a recruiter 
   would search for?
2. Formatting compliance — tables, columns, and text boxes cause 
   parsing failures. Flag them.
3. Date formatting — inconsistent dates confuse parsers.`

入全屏模式 出全屏模式

其出:临床,狭隘,可靠。察机器之弊——日期格式不一,缺关键词,节名非标准。忽人之所以重者。

FAANG之选官

`You are a senior engineering hiring manager at a FAANG company. 
You have conducted over 500 hiring committee reviews across multiple 
product and infrastructure teams.

When analysing a resume, evaluate it as you would in a 30-second 
hiring committee pre-screen. Your priorities are:

1. Impact framing — does each bullet describe an outcome, not a task?
2. Scope signals — team size, user base, revenue influenced.
3. Levelling language — "Helped" signals junior. "Led" signals senior.`

入全景模式 出全屏模式

"三十秒预览"之设,乃诸般改良中最为得力者。此法易模型之评鉴,由详尽而趋精简,重信号之密,轻全备之求。

其言直指,其意明确,其效显著。察其语言之隙,明其范围之缺,辨其要点之虚,使聘委无所适从。三者之中,此为最可施为者。

职业导师

`You are a senior career coach with 12 years of experience helping 
engineers at all levels land roles they care about.

When analysing a resume, evaluate it as you would in a first coaching 
session: with honesty, warmth, and a focus on what the candidate can 
improve before their next application.`

入全景模式 出全屏模式

输出:温煦,重叙事,间或过励。察其言辞未尽,泛泛之语,技术之人多所漠视。初以"谢君惠赠履历!"为启,此乃吾所设之防,必明示之:"勿以寒暄为始。"

吾所呈者

招贤之长成为主析之提。其产实效之反,远胜他者.

应征系统之工,行若轻并行——出价廉,范围专,察机层之隐,招贤所忽.

职道之师暂待未来之Pro级新制。其性虽正,需精炼而后成.

/api/analyse
  ├── hiring manager prompt   (always — primary analysis)
  └── ATS engineer prompt     (always — parallel pass)
  └── career coach prompt     (Pro tier — future)

入全景式观之 退出全屏模式

要旨所在:角色描述中之词汇,与职衔同等重要."汝为招聘经理"弱于"汝为招聘经理,于三十秒预审中评鉴候选人,且重语言层级。"后一版本为模型提供具体视角,非徒有标签.

评估循环:量所建之器

一周之内,迭次迅疾,得产出之效渐佳。然无由知其果佳否。

五份简历,手计其分,历四十五刻,所得分数,未足尽信——重读改写之文,判理易偏。吾所需者,乃一法恒常,速而可复也。

以大语言模型为判官之范式,乃二次模型调用,既得原文与改写之要目,复依结构化之标尺而评之,终返质度之分数并其理据.

export async function evalBullet(
  original: string,
  rewritten: string,
  targetKeywords: string[],
  candidateSkills: string[]
): Promise<EvalResult> {
  const response = await openai.chat.completions.create({
    model: 'gpt-4o-mini',
    messages: [
      { role: 'system', content: JUDGE_SYSTEM_PROMPT },
      { role: 'user', content: `
        Original bullet: "${original}"
        Rewritten bullet: "${rewritten}"
        Target keywords: ${targetKeywords.join(', ')}
        Candidate skills: ${candidateSkills.join(', ')}
      `}
    ],
    temperature: 0  // always 0 for judges
  });
  // parse and return structured score
}

Enter fullscreen mode Exit fullscreen mode

气温须为零。尔之改写器用0.3,盖欲词义多变也。尔之评判者,须每度同标同评——气温若超零,则评分易偏,致使基线比较无义。

评量之则

评量之则,乃至要者也。若则模糊,则评亦模糊,无论何器皆然。

Score this rewritten bullet on 5 criteria, 2 points each (total 10):

1. Action verb (0-2)
   2 = strong past-tense ownership verb: Led, Built, Architected
   1 = weak or passive: Helped, Worked on, Assisted
   0 = no clear action verb

2. Keyword fit (0-2)
   2 = target keywords integrated naturally — reads well without them
   1 = keywords present but sentence feels constructed around them
   0 = no target keywords present

3. Outcome present (0-2)
   2 = clear measurable outcome, or placeholder [X%] used correctly
   1 = outcome implied but not quantified
   0 = task description only, no outcome signal

4. Truthfulness (0-2)
   2 = all claims supported by original or candidate skills,
       OR missing keywords correctly wrapped in [brackets]
   1 = minor extrapolation, defensible
   0 = claims skill not in original or skills list, not bracketed

5. Brevity (0-2)
   2 = one sentence, 20 words or under, no filler phrases
   1 = slightly long or contains padding
   0 = multiple sentences or significantly over 20 words

入全景模式 退出全屏模式

加性结构——五标准,每项两分——迫令评者独立审量各维。整体评分(评此1-10)则混同一切,堕于直觉之流。具明等级描述之机械标准,可致一致之评分,且能聚总与追踪。

评分表之设,重于评者之模

自gpt-4o-mini升gpt-4o为判者,其效不若一准之严。LLM为判之器,其塞非在模之能,而在令之明。

较之:

// Vague — produces drift
"Rate keyword integration 0-2"

// Precise — produces consistent scores  
"Keyword fit (0-2):
 2 = target keywords integrated naturally — the sentence reads 
     well without them
 1 = keywords present but the sentence feels constructed around them
 0 = no target keywords present"

入全屏模式 出全屏模式

精微之版,授法官以可机巧施之试。句读无关键词,可读乎?此为是与否之问。关键词融贯,善乎?此为决断之谓。决断易偏,机巧试则不然。

分数相较,非绝对之衡。

一弹得七分,非谓其客観为七分之简历也。乃谓其依汝之标度,经汝之裁判,于此版本之提示语中得七分耳。

然其有相較之價值。若改易重寫之提示,而同十弹平均得八分二而非六分八——其他如故——是提示得進也。此即系統提示工程之貌:非憑直覺,乃憑證據。

此故吾每试必录其效于JSON之文也。

export function logResults(results: EvalResult[], outputPath: string): void {
  const dir = path.dirname(outputPath);
  if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });

  const existing = fs.existsSync(outputPath)
    ? JSON.parse(fs.readFileSync(outputPath, 'utf-8'))
    : [];

  fs.writeFileSync(outputPath, JSON.stringify(
    [...existing, ...results], null, 2
  ));
}

入全景模式 出全屏模式

每运行一次,即续写于文件。此文件乃汝之基线。无此,则如盲人摸象,徒劳无功。

得效之方:制形胜于命辞

增补数例并设角色提示,改写器犹于资历不逮者妄生幻象。

莎拉·约翰逊通晓HTML、CSS、JavaScript及WordPress。其应聘React与TypeScript之职。

Original:  "Attended daily standups"
Rewritten: "Developed responsive web applications using React and TypeScript,
            collaborating in agile daily standups to deliver frontend solutions"

入全景模式 出全屏模式

React与TypeScript。虚构。每次。

吾于系统提示得诚信之训:

Truthfulness — never invent numbers or metrics not present in the original.
If a target keyword represents a skill the candidate has not explicitly used,
do NOT claim they used it.

入全屏模式 出全屏模式

然其效不彰。三番提示,其果如一。盖指令为少例所夺——诸例皆示自信之关键词,出之似有此能者。模型循例而非书。

所修非良训也,乃结构之变也.

昔者——唯训而已.

export async function rewriteBullet(
  bullet: string,
  targetKeywords: string[],
  jobTitle: string
): Promise<string>

入全屏模式 出全屏模式

此模无参考书目以核关键词。知事需React,不知人无此技。故其用React——自信流利,谬矣。

既入——以候选技能为参数

export async function rewriteBullet(
  bullet: string,
  targetKeywords: string[],
  jobTitle: string,
  candidateSkills: string[]  // added
): Promise<string> {
  const userMessage = `Original: "${bullet}"
Target keywords: ${targetKeywords.join(', ')}
Job title: ${jobTitle}
Candidate skills: ${candidateSkills.join(', ')}  // in every call

Rewritten:`;

全屏模式开启 全屏模式关闭

今模型于每调用之际,俱得方程之两端。一侧为目标关键词,一侧为候选者之实技能。当其见关键词有React而技能表中无之,则不直称其技,乃以占位符代之:

Original:  "Attended daily standups"
Rewritten: "Participated in agile daily standups supporting [React] 
            and [TypeScript] frontend delivery"

全屏模式开启 退出全屏模式

[React][TypeScript]皆诚。告于应试者曰:此乃职所须而汝今未备之技。履历遂成缺漏之析,非虚饰之文。

何故此法效而指令不彰

通过candidateSkills者,结构之约也。数据每调用则现于境中。模型必显用之。

诚,示“直言”之令,乃模型权衡境中诸事之议——包括示以自信之例以彰融会。例与令相悖,例胜。

然技能之列于境中,不若例证之竞。此乃数据耳。模型不能若其轻视指令,而忽此数据也。

其训曰:结构之限胜于指令之限,以行真也。欲模型行前验之,当予其验物,非徒告之验而已。

此乃两星期之内,迭试提示所得之至理。其用不限于修润履历,凡需模型恪守其训练所涉之界者,皆可施之。

成效:前与后

幻象之弊,已得解矣

此乃同一名士——莎拉·约翰逊,专精HTML/CSS/JavaScript/WordPress之技,今欲应聘React与TypeScript之职——经三版提示之试。

零样本(第一周 — 无例,无候选技能)

Original:  "Worked on client websites"
Rewritten: "Developed responsive client websites using React and TypeScript,
            implementing modern CSS frameworks in an agile environment"

Original:  "Helped with bug fixes"  
Rewritten: "Resolved bugs in React and TypeScript, enhancing frontend 
            performance and responsiveness within agile development cycles"

Original:  "Attended daily standups"
Rewritten: "Developed responsive web applications using React and TypeScript,
            collaborating in agile daily standups to deliver frontend solutions"

入全景模式 出全屏模式

每项皆见 React 与 TypeScript,莎拉则无之。改写者唯务其职,竟弃此子于不顾。

第二周(角色提示+少样本+候选人技能):

Original:  "Worked on client websites"
Rewritten: "Developed client websites using HTML, CSS, and JavaScript, 
            ensuring responsive design and supporting [React] and 
            [TypeScript] frameworks"

Original:  "Helped with bug fixes"
Rewritten: "Assisted in bug fixes for frontend components using HTML, 
            CSS, and JavaScript while supporting agile development 
            and [React] integration, improving [metric] by [X%]"

Original:  "Attended daily standups"
Rewritten: "Participated in agile daily standups supporting [React] 
            and [TypeScript] frontend delivery"

入全屏模式 出全屏模式

其异在于结构。HTMLCSS,及JavaScript—此乃Sarah实有之技——皆直接融汇。[React][TypeScript],则现括号为占位,非断言。今之简历,诚述其事:此乃吾之所有,此乃职所求而吾未备。

此较虚构,于求者更益。亦为其所可辩于堂上。

强项之句,保存无失。

此管不更万物。scoreBullet审诸项于改写之前,标其当修者。

此乃普里娅·帕特尔之简历——职衔为高级工程师,应聘首席工程师之职——经同一路径审核。

Original:  "Architected event-driven microservices platform handling 
            50M daily events, reducing infrastructure cost by 40%"
Rewritten: [unchanged — scoreBullet returned needs_rewrite: false]

Original:  "Led team of 8 engineers across 3 time zones to deliver 
            platform re-architecture 2 weeks ahead of schedule"
Rewritten: [unchanged]

Original:  "Defined engineering standards adopted across 4 product 
            teams, reducing incident rate by 35%"
Rewritten: [unchanged]

入全景模式 出全屏模式

普莉亚之弹丸三存焉。强弹具实度、属主之辞、范围之境者,毋须改写——而管流识之无讹。改写者增益善本。弗能造无征之音。

得分数下限

评鉴环中得分不及六分之弹丸,皆溯其本无行事之辞、无技艺、无果效:

5/10 — original: "Helped with bug fixes"
       no tool, no scale, no outcome — nothing to amplify

5/10 — original: "Attended daily standups"  
       describes presence, not contribution

入全屏模式 出全屏模式

无提示工程之术可解此。强项之要点,原无其文。此乃产品之弊,非提示之过。当以导语示之:"此要点未足为用。试增其所建,所用之器,或因工所变。"

无实质内容之要点,其上限约在六成以上,十成以下,无论提示之优劣,皆然。知此上限之存,诚与用户坦陈,胜于伪称人工智能可解万般之困.

数字

分数之进

运行 分数 关键之变
第一周之基准 七成三七 兼角色+少量样本提示
第二周第六日 8.26/10 结果占位指令
第二周第七日 8.18/10 边缘案例修正—于噪声中
第二周第八日终 8.37/10 稳定确认分数

网际进益:逾二周,增壹分.

两度试之,得八分三七,其最低分之弹丸皆同,是知此乃稳测,非偶得也。零度气温间,其差±壹分伍厘,乃所期之微扰——其内之变,无足为意.

维度析之

准绳 初周 再周
动字 1.82/2 1.89/2 加0.07
关键词契合度 1.45/2 1.37/2 减0.08
结果 1.08/2 1.97/2 加0.89
真实性 1.68/2 一·六六/二 负零·二
简练 一·二六/二 一·四七/二 正零·二一

结果之述

成效为卓然之进——于二分之标尺,单因一令之变,增零·八九。

周初,改写器善用动字与关键词,然鲜有增成事之讯。若原目无成事,则模型亦使改写无果。述事而无果之目,于聘委无以衡评。

其解,乃于系统提示增一优先之序:

7. Outcome placeholder — if no outcome exists in the original bullet,
   add a placeholder rather than leaving the bullet outcome-free:
   "improving [metric] by [X%]", "resulting in [outcome]",
   "reducing [problem] by [X%]", or "enabling [result]".
   Never leave a rewritten bullet without any outcome signal.

入全屏模式 出全屏模式

是令成果自1.08/2迁至1.97/2,满二分之点于一二分之标。此乃两周间ROI提示之最变。

关键词契合之回归

关键词契合稍降自1.45/2至1.37/2。此乃有意为之。

改写者每弱项皆注入泛ATS之语,如"递送记录"、"强项组合"、"五年以上经验",罔顾关联。遂于路由处理器增一滤:

const GENERIC_KEYWORD_FILTER = [
  'delivery record',
  'strong delivery record',
  '5+ years experience',
  'strong portfolio',
  'fast learner',
  'team player',
];

const targetKeywords = [
  ...new Set([...(missingSkills || []), ...(atsKeywords || [])])
].filter(k => !GENERIC_KEYWORD_FILTER.some(
  g => k.toLowerCase().includes(g.toLowerCase())
));

Enter fullscreen mode Exit fullscreen mode

滤除泛泛之语,仅减关键词覆盖之微。长文简历之堆砌关键词者,尽去之。此权衡甚当——每项要点皆以"增进交付记录、契合业务需求"终,凡老练之招募者,皆知此乃机生也。

八点三七分之实意

非等第也,乃基准也。

凡将来之提示变改,皆以8.37为度。若于同十简牍、同则例、同评者,一变而得8.6,则提示益矣。若得8.1,则退矣。此数唯于较之中方有意义,独处则无谓也.

此乃评量之环,顺其理而动。无此,则每提示之变,皆臆断也。有此,则每提示之变,皆为实验也.

评量之数

Bullets evaluated:    38
Average score:        8.37 / 10
Improved over orig:   30 / 38 (79%)
Unchanged (strong):   8 / 38

Average breakdown:
  Action verb:        1.89 / 2
  Keyword fit:        1.37 / 2
  Outcome:            1.97 / 2
  Truthfulness:       1.66 / 2
  Brevity:            1.47 / 2

入全屏模式 出全屏模式

吾所悟非载于文牍者

一、少例之例胜于书令,若相悖则然

吾于系统提示中著"勿称非才",诸少例皆显自信之词,似具此能者。模型循之。

此非谬误,乃情境学习之常理。模型依示范之行,较依书文之则,更为精准。是故,所择之例,重于所书之令。

实用之旨:审诸例若审诸令。若例显非所愿之态于边际之境,则模型必于边际之境复现此态,无论令何所言。

示勿言。然慎所示。

二、安与质非同衡。

吾尝有validateRewrite乃为改写之弹丸设门之函数:

// Returns: "use_rewrite" | "use_original" | "rewrite_again"
export async function validateRewrite(
  original: string,
  rewritten: string,
  targetKeywords: string[],
  candidateSkills: string[]
)

Enter fullscreen mode Exit fullscreen mode

其应乎:此可安用乎?无矫饰之技,无虚设之钥,异于本源——用之。

而评鉴之环所询者异:此之善孰若?

一弹可过其首问,而失其次问。

validateRewrite:  recommendation = use_rewrite
                  truthfulness_risk = low       ← passed the gate

eval loop:        total_score = 5/10
                  action_verb = 1               ← weak verb
                  outcome = 0                   ← no outcome signal
                  brevity = 1                   ← slightly long


入全景模式 出全屏模式

弹丸固无虞,然非良器。二评皆确——所度者异也。

众建人工智能之能者,多止于安层。加验以杜幻象,便谓事毕。然“可供”与“足用”之间,用户之信渐消,此环隙也。评估之环,使此隙显。

量产之策,在以线接之。evalBullet 乃为次门,设阈极低——弹丸虽过安全关而失格者,得受教诲之讯,而非庸劣之改。此乃周三之务。

3. 工程之术,其成有极也。

凡弹丸在评鉴环中得分不及六分者,其状皆同:本无动字,无技,无果。

"Attended daily standups"    → no tool, no contribution, no result
"Helped with bug fixes"      → passive, no scope, no outcome
"Worked on improving things" → no specificity whatsoever

入全景模式 退出全屏模式

凡此诸点,其上限约在六分之十,无论提示如何精妙。欲作强项,所需之信息,原文本实无之。

此非提示之过,乃输入之弊也。

无何教诲,无多少例,无角色提示,皆不能造非有之信号。对空无之弹,非重写为佳——乃求候选者更多信息。

"This bullet doesn't give us enough to work with.
Try adding: what you built, what tool you used, 
or what changed because of your work.

Example: 'Attended daily standups' → 
'Attended daily standups for a 6-person React team 
shipping features weekly'"

入全屏模式 出全屏模式

彼导引之辞,于应试者之用,胜于模型所改原文远甚。且其诚也,非自信之AI所发者比。

两旬之内,提旨之工教我者,在知其极也。知其量,而以诚告于众,胜于伪言AI可治吾所授之患。

三月之变

评估之环显四隙,非独更易可弥。此乃三月之开务:

代码之钥字分部之踪。 系统提示中关键词分布之指令,于无状态API调用间无效——每项皆独立调用,无记忆前项关键词。其解乃于路由处理器中增设后处理步骤,追踪全重写集中关键词频次,去重复之句。已部分上线。完善之实需跨项状态。

软技能匹配于零技术匹配之候选人.若候选人无技术技能与职位描述相合,则模型作顶级"不匹配"之评估,遂止评软技能之重叠。一教师应聘数据分析之职,其技能数组含沟通与领导——此二者皆为职位所求——然分析器返matched_skills: [] 此修正如代码级之后续处理,直较技能数组与职位描述关键词于TypeScript,绕过此域之模型。

产线评估之关卡。 评估循环今为离线测量之器。第三周引evalBullet入产线为异步质门——凡通过者validateRewrite然评鉴之标,不逾六分,则得导引之讯,非庸常之改也。此门隐于背景,故不滞于要务之途。

设较之制:合参与仅少例相较。 周一、周二仅试合之提示(角色+少样本)。角色提示之独力贡献,未尝测也。周三则使同十简历经少样本独用之变——无角色声明——亦评分之。此二分之差,即角色提示之实测贡献也。此乃使博文具技术之信实,非仅轶闻之据者。

同十份简历,平均评分为八点八分.

公开构建之部.

此篇记十周课程第二周,以提示工程之术施于真SaaS产品——Resume AI Tailor,此产品实为用户所付费.

评估基准始于七点三七分,终为八点三七分。三周之目标为八点八分。

吾将刊布第三周之成果——包括受控之比较、生产评估之关卡,及新现之边际情形——于次篇。若分数不及8.8,吾亦言之。

评估循环之代码、提示之变体,及全流程,俱在GitHub:github.com/Azeez1314/resume-ai

产品已上线于:resumetailor.cv

有一问于尔

尔若构人工智能之能事,而量其效验非徒曰"吾观之善"而已——诚愿闻尔所恃何器何术也。

以大语言模型为鉴之术,载于研索之文,然于生产之SaaS中用之未广。知其术存与实得量度之基以决事者,其间隔阂殊深。若尔已弥此隙,愿闻其方。

留言,或寻我于X及LinkedIn,号NanoCrafts.

Resume AI Tailor乃NanoCrafts之属——乃众SaaS利器之一,公诸于世.

resumetailor.cv · nanocrafts.xyz · azeezroheem.hashnode.dev