Nitin S Dixit

[𝐉𝐨𝐝 𝐈𝐧𝐬𝐚𝐧𝐞 @jod_insane]

RETWEET IF YOU AGREE

[Swaksh Tripathi @SwakshTripathi]

I support PM Narendra Modi 💯

Today, Tomorrow, Forever

[Nitin S Dixit @dixycola]

@ANI Why is he wearing shoes only in one leg?

[Nitin S Dixit @dixycola]

@yajnshri purvashadha

[Nitin S Dixit @dixycola]

@Rg197923 @srinipartha1 Because Sri Chaitanya Mahaprabhu, who is Krishna himself, asked to do so.

[Nitin S Dixit @dixycola]

@DocPriyamMD So when does the body start burning fat? After ending glucagon reserves?

[Chota Don @choga_don]

This kid is literally becoming a nightmare for left Liberals 🔥

After cooking Dhruv Rathee, he is BASHING Cockroach Janata Party & its founders brilliantly.

Liberals tried to silence him by copyright strike but now he’s ROASTING them at 10x speed now.

More power to him 🙌🔥

[सूत्रधार - भारत की कहानियाँ @MySutradhar]

Ashvatthama's Night Raid

Ashvatthama, accompanied by Kripa and Kritavarma, reached the enemy camp but found its entrance guarded by a terrifying, divine being of immense power and blazing form. Undeterred, Ashvatthama attacked with celestial weapons, darts, sword, and mace, but the being effortlessly devoured or nullified all his weapons, rendering his efforts futile.

[madhu trehan @madhutrehan]

Guess how many western music artists stole from Asha Bhonsle? Check this out!

[Nitin S Dixit @dixycola]

@Am_Shai Unfortunately, it failed. Generally, what is the failure rate of IS30s and IB30s?

[Nitin S Dixit @dixycola]

It's confirmed @aravind is the time traveller

[Nitin S Dixit @dixycola]

@SwayamTewari 0 pp

[FalconUpdatesHQ @FalconUpdatesHQ]

This needs to go viral 🔥#महाशिवरात्रि

[سماح | 🧚🏾‍♀️samah @samah_fadil]

the oprah episode that vanished: on february 17, 1988, oprah aired a “satanism discussion” episode where a woman revealed the details of what her family was up to behind closed doors. what do you notice?

[Alok Bhatt @alok_bhatt]

Few young Hindu men discovered an ancient Hindu temple in Rajsamand and they are determined to restore- help them folks

[Nitin S Dixit @dixycola]

@MayMayspeaks Two

[The Tathya @_TheTathya]

Abdul Misled Tourists for 50 Years at Ellora’s Kailash Temple… ⚠️😱

> A traveller hires the only “official” guide available... Abdul.
> Inside the temple, Abdul starts narrating shocking fabrications.
> “Shiva doubted Parvati.”
> “Women fall for other men easily.”
> “Shiva came as a beggar to test her.”
> A 19-year-old boy hears it… and instantly believes it.
> The traveller realizes lakhs may have heard the same lies over decades.
> When questioned, Abdul explodes: “Nikal jaa idhar se!”
> Another tourist corrects him... Abdul throws him out too.
> This man has been guiding here for 50 years.
> And ASI still calls him “government-approved”.

> The traveller records everything.
> Goes straight to ASI.
> Files a complaint with video proof.
> ASI says: “We’ll forward it… no guarantee of action.”

50 years of misinformation.
Still no accountability.

This isn’t about religion.
It’s about stopping distortion of sacred history.

If one Abdul could rewrite Shiva–Parvati’s story so easily… imagine how many more are doing the same across India.

Please Retweet and create awareness... 🚨🔥

@ASIGoI

[Nitin S Dixit @dixycola]

@Mrcryptoxwhale Done

[Femke Plantinga @femke_plantinga]

Your RAG system is failing?

It's not your vector database - it's how you're chunking your data.

Everyone focuses on picking the perfect vector database or embedding model. But here's what actually makes or breaks your RAG system: how you prepare your data before it ever gets embedded.

The fundamental challenge with chunking is this: your chunks need to be small enough for precise vector search, but complete enough to give your LLM the context it needs to generate accurate responses. Get this wrong, and even the best retrieval system will fail.

𝗖𝗵𝘂𝗻𝗸𝗶𝗻𝗴 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 (𝗳𝗿𝗼𝗺 𝘀𝗶𝗺𝗽𝗹𝗲 𝘁𝗼 𝗮𝗱𝘃𝗮𝗻𝗰𝗲𝗱 ):

𝗙𝗶𝘅𝗲𝗱-𝗦𝗶𝘇𝗲 – Split by token count (~512). Fast and simple, but cuts mid-sentence.

𝗥𝗲𝗰𝘂𝗿𝘀𝗶𝘃𝗲 – Splits by structure (paragraphs → sentences). Best for articles.

𝗗𝗼𝗰𝘂𝗺𝗲𝗻𝘁-𝗕𝗮𝘀𝗲𝗱 – Uses inherent structure (headers, HTML tags). Perfect for structured content.

𝗦𝗲𝗺𝗮𝗻𝘁𝗶𝗰 – Detects topic changes via embeddings. Variable-length chunks, one idea each.

𝗟𝗟𝗠-𝗕𝗮𝘀𝗲𝗱 – Uses AI to identify propositions. High quality, high cost.

𝗔𝗴𝗲𝗻𝘁𝗶𝗰 – AI agent decides which strategy to use per document. Custom but expensive.

𝗟𝗮𝘁𝗲 𝗖𝗵𝘂𝗻𝗸𝗶𝗻𝗴 – Embeds full document first, then derives chunks. Retains full document context.

𝗛𝗶𝗲𝗿𝗮𝗿𝗰𝗵𝗶𝗰𝗮𝗹 – Multiple layers at different detail levels. Start broad, drill down when needed.

𝗔𝗱𝗮𝗽𝘁𝗶𝘃𝗲 – Dynamic chunk sizes based on content density. Small for complex sections, larger for simple ones.

𝗕𝘂𝘁 𝗵𝗲𝗿𝗲'𝘀 𝘄𝗵𝗮𝘁 𝗺𝗼𝘀𝘁 𝗽𝗲𝗼𝗽𝗹𝗲 𝗺𝗶𝘀𝘀:
Chunking isn't a standalone problem. It's one piece of 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 – the architecture that determines what information your LLM sees and when.

You also need query augmentation, agents, memory, and tools working together. You can't optimize chunking in isolation.

We just released a guide covering all of this, including decision trees to help you pick the right chunking strategy for your use case.

Get your free copy here → https://t.co/FRPyeDeyPh

[The Jaipur Dialogues @JaipurDialogues]

This Bomb Blast at Delhi is not an Internal Security Issue

But it is an Islamic Problem!

RT if you Agree!