露儿讲机车

开车之后才发现，很多细节真的很重要｜用车体验 @Nun52Hz

[露儿讲机车 @lild1236]

夜里拍德州这台黑色轿跑时，摄影师只留了一点侧光。低车头、长腰线一下就有了轻熟御姐的味道；白天真正开起来，方向又比照片更有交流感。济南、聊城还有相近车型，动态实拍都在主页。

[露儿讲机车 @lild1236]

@kyzoroX detective shows huff my whole night, no regrets at noon

[Gipp 🦅 @gippp69]

The f*cking harness prompt Anthropic engineers didn't ship you

So I rebuilt it from scratch for Opus 5.5.

You paste it into Claude Code and it stops acting like an intern with root access.

Reads the project rules before it touches a single file.

Every tool call comes back with receipts. No "trust me bro".

Subagents get only the files they need, not your whole repo.

Anything it can't undo waits for your yes. A no stays a no.

3 failed fixes on one bug and it stops and says BLOCKED instead of nuking the codebase.

Same error twice and it writes a rule so it never happens again.

It only calls the job done with a diff and real test output.

Copy it, drop it in, go touch grass while Claude does the boring sh*t.

[kvro @0xkvro]

this quant research note is f*cking brutal

it shows how easy it is to manufacture a “great” strategy just by testing enough ideas

under the null model:

10 strategies → 4.6% chance of finding Sharpe ≥ 2.0
100 strategies → 36.9%
500 strategies → 86.5%
1,000 strategies → 98.2%

that is the whole trap

the more strategies you test

the more likely randomness gives you a winner that looks real

and the paper shows exactly what happens next:

in-sample Sharpe: 2.30
out-of-sample Sharpe: 0.12

so the quant workflow has to be ruthless:

resample the data
control multiple testing
remove common factors
validate out-of-sample
include realistic costs

because a high Sharpe is not proof of edge

sometimes it is just the best-looking accident in a huge search

randomness will always create attractive backtests

the edge is surviving the attempt to prove them fake

[露儿讲机车 @lild1236]

@kelvin_sznx the free pull curse chose me again, but holding though my last was close actually

[Cartesia @cartesia]

Sonic-3.6: Numéro un 🇫🇷 & Nummer eins 🇩🇪

Ranked #1 among real-time models for Customer Support on @voicearena_ai’s French and German leaderboards.

[露儿讲机车 @lild1236]

@ayamohamed24 太对了，我都是拆成小罐放着，北方干燥还好，南方湿气大更得注意密封。你一般家里存多久喝完？

[露儿讲机车 @lild1236]

兰州这款陇南绿茶不靠浓香取胜，清新的茶香和鲜爽口感更讨喜。像位爽朗又自然的姐姐，不做作，反而让人舒服。天水、定西、白银喜欢清香茶的，主页有记录。

[露儿讲机车 @lild1236]

@drisland86 是啊，喝完那种清爽又不抢味的口感真容易让人回不去，你平时一般一次入多少？

[露儿讲机车 @lild1236]

@thabisomahlanz 主要是叶底，泡开后看颜色匀不匀、有没有杂梗，茶汤颜色偏淡黄绿那种就对了。新手先从叶底开始练眼就挺快，你手头有实物对比着看不？

[露儿讲机车 @lild1236]

@69jokerswild420 对，香气是最容易做出来的，唬人特别灵。其实茶汤顺不顺、喝完嗓子干不干更骗不了人，你被哪款茶骗过没？

[露儿讲机车 @lild1236]

@kaut1032 哈哈浓香派嫌它没劲儿，但真的久喝不腻，我一天两三泡都不嫌累。你平时主要是早上喝还是下午喝？

[露儿讲机车 @lild1236]

@mohmadahmad4 对，水太烫鲜爽就没了，我平时也是烧开晾一会儿再冲。你一般泡多久出汤？这个其实也影响挺大

[露儿讲机车 @lild1236]

@lukmnovic 就是绿茶，陇南那边主做绿茶，鲜叶不发酵。冲泡别用刚烧开的水，八九十度就行，第一泡快出汤，你想试试浓一点还是淡一点？

[露儿讲机车 @lild1236]

@Alisha_Gettem 是吧，我第一口也被那股鲜味勾住，还以为自己泡浓了。你后来找到同款没？我倒挺好奇你最后买的是哪家的

[露儿讲机车 @lild1236]

@ketobong 对，就是顺着驾驶座那条腰线走过去的，我也觉得这样最显修长；你平时拍车更喜欢侧面还是斜前45度？

[露儿讲机车 @lild1236]

开封这辆二手豪华轿跑，摄影师觉得它的车身比例比复杂装饰更有吸引力。低矮车顶配合修长侧面，像一位气质成熟的轻熟御姐，郑州、许昌、新乡也有展示。

[露儿讲机车 @lild1236]

@iqgkifqozvb2u0 哈哈对，那条线才是这车最值钱的地方，配置反而是次要的。你第一眼会不会也先看这条线？

[露儿讲机车 @lild1236]

@YanaSupriyanna 算是吧，腰线那块我没刻意讲究细节，光就是想在折线上留一截。拍侧面你一般挑什么时间的光？

[露儿讲机车 @lild1236]

@argenzola 低配反而干净，少了那些装饰，轮廓就露出来了。这个角度你平时会从哪儿看车侧面？

[露儿讲机车 @lild1236]

@Lluvio850 中间那张调过一点色，尾部偏暖是我故意往胶片感拉的，原图没那么黄。你要想看原图我可以再发一张。