aiweb3rk

[aiweb3rk @aiweb3rk]

this is insane lecture .

[venus @RitOnchain]

OpenAI Research Scientist just revealed the scaling playbook that makes 99% of current AI architectures obsolete.

36-minutes. free. By MIT PhD & OpenAI dev team.

"What is better in the long term almost necessarily looks worse now."

the macro view for quants it covers:
•why adding custom mathematical structures bottlenecks your model at scale
•the hidden engineering flaw of bidirectional attention in multi-turn chat
•complete architectural breakdown: encoder-decoder vs decoder-only
•why raw compute scaling destroys hyper-specific inductive biases

Watch full video then read article below.

[venus @RitOnchain]

OpenAI Research Scientist just revealed the scaling playbook that makes 99% of current AI architectures obsolete.

36-minutes. free. By MIT PhD & OpenAI dev team.

"What is better in the long term almost necessarily looks worse now."

the macro view for quants it covers:
•why adding custom mathematical structures bottlenecks your model at scale
•the hidden engineering flaw of bidirectional attention in multi-turn chat
•complete architectural breakdown: encoder-decoder vs decoder-only
•why raw compute scaling destroys hyper-specific inductive biases

Watch full video then read article below.

[aiweb3rk @aiweb3rk]

@RitOnchain Bookmarked for tonight.

[aiweb3rk @aiweb3rk]

@RitOnchain this is really insane article man.

[aiweb3rk @aiweb3rk]

rag mining . Wow.

[venus @RitOnchain]

https://t.co/fbXcG5WMYo

[venus @RitOnchain]

https://t.co/fbXcG5WMYo

[aiweb3rk @aiweb3rk]

@RohOnChain insane man.

[aiweb3rk @aiweb3rk]

this is really insane

[Roan @RohOnChain]

this is f*cking dangerous

build hedge fund using "loop engineering" that prints alpha 24/7 (Full Guide)

if I had this a year ago, I would've built my hedge fund in a week instead of a year

bookmark before someone takes it down

[Roan @RohOnChain]

this is f*cking dangerous

build hedge fund using "loop engineering" that prints alpha 24/7 (Full Guide)

if I had this a year ago, I would've built my hedge fund in a week instead of a year

bookmark before someone takes it down

[aiweb3rk @aiweb3rk]

Must watch video

[Roan @RohOnChain]

My friend applied to 100+ quantitative researcher jobs in two years. No PhD. No callbacks.

Few days ago Jane Street offered him $600,000.

The thing that helped him the most was this master class by IMC, one of the biggest HFT market makers in the world. Free on YouTube. One hour.

Their head of alpha research explains how top trading firms actually build their models. Not the YouTube tutorial version. The real one.

He watched it on a Sunday. Paused it nine times. After that hour he told me something I didn't believe. "It's embarrassingly simple."

Three days later he applied to Jane Street.

Every single question they asked him, he knew from that video.

[Roan @RohOnChain]

My friend applied to 100+ quantitative researcher jobs in two years. No PhD. No callbacks.

Few days ago Jane Street offered him $600,000.

The thing that helped him the most was this master class by IMC, one of the biggest HFT market makers in the world. Free on YouTube. One hour.

Their head of alpha research explains how top trading firms actually build their models. Not the YouTube tutorial version. The real one.

He watched it on a Sunday. Paused it nine times. After that hour he told me something I didn't believe. "It's embarrassingly simple."

Three days later he applied to Jane Street.

Every single question they asked him, he knew from that video.

[ritesh.hft @hyperkquant]

DL must watch by lex!

[ritesh.hft @hyperkquant]

insane optimization paper.

[aiweb3rk @aiweb3rk]

Must read

[Roan @RohOnChain]

https://t.co/JRsWSeUfrN

[Roan @RohOnChain]

https://t.co/JRsWSeUfrN

[aiweb3rk @aiweb3rk]

Must watch

[venus @RitOnchain]

most traders build strategies assuming tomorrow will look exactly like yesterday. it won't.

51-minutes. free. by a senior quant systems architect.

here's what it covers:
•the 3-equation math behind hidden markov models
•python implementation for soft-switching regime signals
•factor rotation setups for bull, bear, and neutral markets
•19.41% annualized return vs 10.80% spy buy-and-hold benchmark
•the 4 core failure modes

bookmark & watch today. then read the article below.

[venus @RitOnchain]

most traders build strategies assuming tomorrow will look exactly like yesterday. it won't.

51-minutes. free. by a senior quant systems architect.

here's what it covers:
•the 3-equation math behind hidden markov models
•python implementation for soft-switching regime signals
•factor rotation setups for bull, bear, and neutral markets
•19.41% annualized return vs 10.80% spy buy-and-hold benchmark
•the 4 core failure modes

bookmark & watch today. then read the article below.

[aiweb3rk @aiweb3rk]

@RitOnchain crazy paper

[aiweb3rk @aiweb3rk]

must read paper for quants .

[venus @RitOnchain]

A 10bps alpha signal is worthless if execution costs 12bps.

This paper uses reinforcement learning to optimize execution directly from order book dynamics.

The edge isn't finding alpha. It's keeping it.

Bookmark and Read the article about DQN Quant Framework.

[venus @RitOnchain]

A 10bps alpha signal is worthless if execution costs 12bps.

This paper uses reinforcement learning to optimize execution directly from order book dynamics.

The edge isn't finding alpha. It's keeping it.

Bookmark and Read the article about DQN Quant Framework.