Sharad Varshney

Principal - Agentic AI

[Sharad Varshney @varshnes]

@radd280 @ripster47 https://t.co/imtkaGNB6M

@ripster47 Is this for real??? Looks like some fake accoun trying to run pump n dump.

[Sharad Varshney @varshnes]

@radd280 @ripster47 @ripster47 is this your account or some fake company trying to run Pump n dump here

[Sharad Varshney @varshnes]

@ripster47 @BlackBoxStocks Sometimes I pike early based on the losses that I had seen in the past but being trading with you for over a year... I have been slowly trusting the system. I need to go through all the concepts again and re-write them and get more conviction.

[Ripster @ripster47]

💡🎁 My Next Holiday Giveaway
Stock & Options Scanner i use everyday

3 ANNUAL PREMIUM ACCOUNTS ($2000 worth each)

👉 https://t.co/QVsvOT8kLQ

✅RT/LIKE THIS , Follow @BlackBoxStocks
✅Tell me what is lacking in your trading & why?

Results on Saturday

[Sharad Varshney @varshnes]

@justpeoplebybg Great Intiative

[Sharad Varshney @varshnes]

@elonmuskTN 15

[Sharad Varshney @varshnes]

@ripster47 Trading 0 DTE is not for everyone, only people with good experience trading options should trade 0 DTE. Trade with @ripster47 , he will make sure you make some money rather than losing BIG.

[Sharad Varshney @varshnes]

@RobinerdGames @sama You bet!

[Sharad Varshney @varshnes]

@tom_doerr what envrionemnt you are using to call llama3.1?

[Guillaume Lample @ NeurIPS 2024 @GuillaumeLample]

Very happy to release our new small model, Mistral NeMo, a 12B model trained in collaboration with @nvidia. Mistral NeMo supports a context window of 128k tokens, comes with a FP8 aligned checkpoint, and performs extremely well on all benchmarks. Check it out!
https://t.co/ODovtWkAKK
https://t.co/Lc37V0rC5w

[Vaibhav (VB) Srivastav @reach_vb]

Probably the craziest week in Open Source AI (yet):

1. Mistral (in collaboration with Nvidia) dropped Apache 2.0 licensed NeMo 12B LLM, better than L3 8B and Gemma 2 9B. Models are multilingual with 128K context and a highly efficient tokenizer - tekken.

2. Apple released DCLM 7B - truly open source LLM, based on OpenELM, trained on 2.5T tokens with 63.72 MMLU (better than Mistral 7B)

3. HF shared SmolLM - 135M, 360M, & 1.7B Smol LMs capable of running directly in the browser; they beat Qwen 1.5B, Phi 1.5B and more. Trained on just 650B tokens.

4. Groq put out Llama 3 8B & 70B tool use & function calling model checkpoints - achieves 90.76% accuracy on Berkely Function Calling Leaderboard (BFCL). Excels at API usage & structured data manipulation!

5. Salesforce released xLAM 1.35B & 7B Large Action Models along with 60K instruction fine-tuning dataset. The 7B model scores 88.24% on BFCL & 2B 78.94%

6. Deepseek changed the game with v2 chat 0628 - The best open LLM on LYMSYS arena right now - 236B parameter model with 21B active parameters. It also excels at coding (rank #3) and arena hard problems (rank #3)

There's a lot more; Arcee (mergekit) released a series of LLMs, each better than the other, and Numina and HF Numina 72B (based on Qwen 2) and Math datasets, Mixbread with embedding models (english + german) and a lot more!

It's fun to see so many releases next week with L3 405B (?) and companions; we might see a shift in the Open LLM landscape! See you next week!

What else did I miss? 🤗

[Sharad Varshney @varshnes]

Air India service still very bad. Don’t even have words

[Sharad Varshney @varshnes]

Air India flight is still having same issues when it was under Govt. Of India… no changes even under Tata… their system automatically cancelled our fully booked and reserved tickets. #stuckinIndia

[Sharad Varshney @varshnes]

@Houxi1314520 It’s Indian sweet dish called Ghevar

[Sharad Varshney @varshnes]

@DhaarmikAi Is this one of the instance or is this true in general that any business which was doing great was pressurized to be sold?

[Sharad Varshney @varshnes]

@POTUS Why India is not in this list? Can anyone help me understand ?

[Sharad Varshney @varshnes]

@PMOIndia @narendramodi Why cant India allow dual citizenship with few democratic countries - atleast Indian citizens do not need to renounced their own passports and get treated foreigners in their own birth country! Please look into this as you move along your 3rd term.

[Dylan Patel @dylan522p]

ChatGPT system prompt is 1700 tokens?!?!?
If you were wondering why ChatGPT is so bad versus 6 months ago, its because of the system prompt.
Look at how garbage this is.
Laziness is literally part of the prompt.
Formatted in the paste bin below.
https://t.co/XSA85dys1I

[Sharad Varshney @varshnes]

LLMs don’t have hallucination problem….

[Andrej Karpathy @karpathy]

# On the "hallucination problem"

I always struggle a bit with I'm asked about the "hallucination problem" in LLMs. Because, in some sense, hallucination is all LLMs do. They are dream machines.

We direct their dreams with prompts. The prompts start the dream, and based on the LLM's hazy recollection of its training documents, most of the time the result goes someplace useful.

It's only when the dreams go into deemed factually incorrect territory that we label it a "hallucination". It looks like a bug, but it's just the LLM doing what it always does.

At the other end of the extreme consider a search engine. It takes the prompt and just returns one of the most similar "training documents" it has in its database, verbatim. You could say that this search engine has a "creativity problem" - it will never respond with something new. An LLM is 100% dreaming and has the hallucination problem. A search engine is 0% dreaming and has the creativity problem.

All that said, I realize that what people *actually* mean is they don't want an LLM Assistant (a product like ChatGPT etc.) to hallucinate. An LLM Assistant is a lot more complex system than just the LLM itself, even if one is at the heart of it. There are many ways to mitigate hallcuinations in these systems - using Retrieval Augmented Generation (RAG) to more strongly anchor the dreams in real data through in-context learning is maybe the most common one. Disagreements between multiple samples, reflection, verification chains. Decoding uncertainty from activations. Tool use. All an active and very interesting areas of research.

TLDR I know I'm being super pedantic but the LLM has no "hallucination problem". Hallucination is not a bug, it is LLM's greatest feature. The LLM Assistant has a hallucination problem, and we should fix it.

</rant> Okay I feel much better now :)