#hackerrank #python #programming #mechanicalengineering #coding #problemsolving #engineeringstudent #continuouslearning #skilldevelopment #careergrowth | JOSHVA G
0:08
#hackerrank #python #programming #mechanicalengineering #coding #problemsolving #engineeringstudent #continuouslearning #skilldevelopment #careergrowth | JOSHVA G
3 weeks ago
linkedin.comJOSHVA G
The big unlock is speed + efficiency.Around 80 tokens per second.But the better part is token usage.On a hard coding test, it completed tasks with far fewer tokens than leading models.Less waste.Less waiting.More output.
8:17
The big unlock is speed + efficiency.Around 80 tokens per second.But the better part is token usage.On a hard coding test, it completed tasks with far fewer tokens than leading models.Less waste.Less waiting.More output.
1.1K views2 weeks ago
x.comJulian Goldie SEO
BSF HCM & ASI Reasoning Class | CISF Fireman Practice Set - 05 | Reasoning For BSF Tradesman
48:09
BSF HCM & ASI Reasoning Class | CISF Fireman Practice Set - 05 | Reasoning For BSF Tradesman
23.3K views3 weeks ago
YouTubeRojgar with Ankit Defence
Thoughts on the Schema harness:TLDR: Without inspecting the code/harness it's tough to tell how much human intelligence was baked in. There are signs it's non-zero, but unconfirmed. Regardless, the team did interesting work and I'm glad they are pushing world modeling ideas forward. Fun way to present their ideas.My notes:> Core idea. The latent world representation is a program, not a vector — so it is interpretable (a text file you can read and diff), verifiable (replayable against recorded re
0:47
Thoughts on the Schema harness:TLDR: Without inspecting the code/harness it's tough to tell how much human intelligence was baked in. There are signs it's non-zero, but unconfirmed. Regardless, the team did interesting work and I'm glad they are pushing world modeling ideas forward. Fun way to present their ideas.My notes:> Core idea. The latent world representation is a program, not a vector — so it is interpretable (a text file you can read and diff), verifiable (replayable against recorded re
583.9K views1 week ago
x.comGreg Kamradt
A food delivery company just beat GPT 5.5 on a real coding test. With a FREE model.Meituan dropped LongCat 2.0 on June 30th. Yes, the dinner delivery people.Here's why it's wild:→ It scored 59.5 vs GPT 5.5's 58.6 on SWE-Bench Pro. Real bugs, not trivia→ 1 million token memory. Most models forget after 100k→ 1.6 trillion parameters, but only 48 billion wake up per task→ Trained on 50,000 Chinese chips. Zero NVIDIA. First time at this size→ The best part: they secretly launched it as "Owl Alpha" a
7:47
A food delivery company just beat GPT 5.5 on a real coding test. With a FREE model.Meituan dropped LongCat 2.0 on June 30th. Yes, the dinner delivery people.Here's why it's wild:→ It scored 59.5 vs GPT 5.5's 58.6 on SWE-Bench Pro. Real bugs, not trivia→ 1 million token memory. Most models forget after 100k→ 1.6 trillion parameters, but only 48 billion wake up per task→ Trained on 50,000 Chinese chips. Zero NVIDIA. First time at this size→ The best part: they secretly launched it as "Owl Alpha" a
1.8K views3 weeks ago
x.comJulian Goldie SEO
three minutes and eight seconds. that's how long grok 4.5 took to build this entire three.js world from my locked spec. one prompt in grok build, zero steering, first full run. seven files, about 2,200 lines. floating island with procedural bark and rock, grass, a day night cycle, a lake with lily pads, fireflies, flying goblins. and a 12 test acceptance gate it had to pass without touching the tests. all green.i benchmark coding agents constantly and the speed still caught me off guard. the tim
4:19
three minutes and eight seconds. that's how long grok 4.5 took to build this entire three.js world from my locked spec. one prompt in grok build, zero steering, first full run. seven files, about 2,200 lines. floating island with procedural bark and rock, grass, a day night cycle, a lake with lily pads, fireflies, flying goblins. and a 12 test acceptance gate it had to pass without touching the tests. all green.i benchmark coding agents constantly and the speed still caught me off guard. the tim
10.4K views2 weeks ago
x.comSudo su
Claude Code用の公式プロンプトライブラリ、Anthropicが無料で公開していたんです。「Claude Codeで何をどう頼めばいいかわからない」という入口の壁が、これでかなり下がりそうです。収録されているカテゴリはこんな感じです。・新規リポジトリ（コードの保管場所）を素早く理解する・特定のロジック（処理のコード）がどこにあるか探す・失敗しているテストを修正する・テストを追加して通す・コミット前に変更内容をレビューする開発の日常シーンをまるっとカバーしていて、コピーして貼るだけで使えます。単なるコピペ集じゃないのがポイントで、各プロンプトに「なぜこれが効くのか」の解説も付いているんです。効くプロンプトには共通のパターンがあって、・結果を指定する、手順は指定しない・「test」「run」「verify」を一緒に頼んで自己チェックさせる・既存のファイルや命名規則を参照させるこの3つが揃うと、Claude Codeが自分で判断・確認しながら動いてくれます。使い込んでいくと「ライブラリ外の場面でも応用できる」感覚が身に付いてくるので、最初はコピペで試しつつ、徐々に自分流にアレンジして
0:33
Claude Code用の公式プロンプトライブラリ、Anthropicが無料で公開していたんです。「Claude Codeで何をどう頼めばいいかわからない」という入口の壁が、これでかなり下がりそうです。収録されているカテゴリはこんな感じです。・新規リポジトリ（コードの保管場所）を素早く理解する・特定のロジック（処理のコード）がどこにあるか探す・失敗しているテストを修正する・テストを追加して通す・コミット前に変更内容をレビューする開発の日常シーンをまるっとカバーしていて、コピーして貼るだけで使えます。単なるコピペ集じゃないのがポイントで、各プロンプトに「なぜこれが効くのか」の解説も付いているんです。効くプロンプトには共通のパターンがあって、・結果を指定する、手順は指定しない・「test」「run」「verify」を一緒に頼んで自己チェックさせる・既存のファイルや命名規則を参照させるこの3つが揃うと、Claude Codeが自分で判断・確認しながら動いてくれます。使い込んでいくと「ライブラリ外の場面でも応用できる」感覚が身に付いてくるので、最初はコピペで試しつつ、徐々に自分流にアレンジして
134 views2 weeks ago
x.comクロニキ｜Claude Code専門家
0:39
GROK 4.5 RANKS #1 FOR REAL-WORLD CODING QUALITYIn a direct head-to-head test by Command Code, Grok 4.5, Fable 5, and GPT 5.5 were given the exact same prompt to build a game.Results:• Grok 4.5: 9/10 — Produced a genuinely polished, finished mobile game experience.• Fable 5: 7.5/10• GPT 5.5: 7/10The other two models felt rushed and lacked refinement, while Grok 4.5 delivered clean, high-quality output.This isn’t just another benchmark number — it’s a real-world test showing Grok 4.5’s strength in
9.8K views2 weeks ago
x.comTesla Owners Silicon Valley
3:26
终于等到了！就在刚刚，@OpenAI 一口气把GPT-5.6「全家桶」——Sol、Terra、Luna全放了出来。毫不夸张地说，Sol这次就是冲着Claude Fable 5来的：1. Agents' Last Exam：Sol得分53.6%，领先Fable5 13.1分；2. Terminal‑Bench 2.1：Sol常规版本88.8%，开启Ultra模式高达91.9%，大幅超过Fable5的83.1%；3. Coding Agent Index：Sol拿到80分刷新行业SOTA，领先对手2.8分，代码能力再创新高。更刺激的的价格——Fable 5每百万token的输入是$10、输出是$50；Sol直接对折，只要$5和$30，Terra再砍半到$2.5和$15，Luna干脆只要$1同在今天，ChatGPT和Codex终于完成合体，OpenAI「超级应用」真的来了。未来ChatGPT将成为唯一入口，告别多个软件来回切换。AI行业新一轮内卷正式开启。
2.2M views2 weeks ago
x.com加密Krystal
1:40
GM GN距离2027年仅剩178天polo的AI日报 7月10日，25条精选 ⬇️(今天确实是炸裂的一天。OpenAI 一口气发了 GPT-5.6 系列和ChatGPT Work,英伟达被法国盯上了)1. 马斯克公开称赞Anthropic是当前AI领域的领导者，承诺不切断算力。Anthropic每月花12.5亿美元买下Colossus 1数据中心全部300兆瓦算力，合同签到2029年，总额约400亿美元。💡能花钱解决的问题都不是问题2. OpenAI发布GPT-5.6系列，Sol、Terra、Luna三款定位分明。Sol在Agents Last Exam上拿53.6分，比Claude Fable 5高13.1分，Coding Agent Index直接干到80分，Fable 5仅2.8分。💡Sol把Fable甩了3. OpenAI推出ChatGPT Work，由GPT-5.6驱动的跨应用自主智能体，能拆解复杂项目、跨应用收集信息、连续工作数小时，内置Codex技术，每周超500万用户使用。💡AI从对话助手进化成远程实习生4. GPT-5.6成为Microsoft 365 Cop
3.6K views2 weeks ago
x.comPolo贱🕊️买美股上币安
3:48
Black democrats want reparations. Look at your own party, they’re the ones that took it from you. Let’s review what the true story of what the Democratic Party has really done to Black people:• 1850: Democrats passed Fugitive Slave Act forcing Northerners to recapture escaped slaves.
• 1854: Democrats passed Kansas-Nebraska Act expanding slavery into new territories.
• 1857: Democrats celebrated Dred Scott decision denying Black citizenship.
• 1860: Southern Democrats seceded to preserve slavery
26.1K views1 week ago
x.comSovey
See more
Static thumbnail place holder