←
sysls5,525 posts
sysls avatar

sysls

@systematicls

All in @openforage. I thrived in all of the largest hedge funds managing systematic investment processes.

◷ Joined April 2024▧ 361 Photos and videos
sysls avatar
sysls@systematicls·Nov 15

I think the explosion of “neolabs” in quantitative trading has been brought about by a misconception either in the competitive landscape OR where money is made in quantitative trading. The idea that a startup claims to be the first to pursue “the bitter lesson” in quantitative trading is absolutely ludicrous. Top tier quant funds have long been hiring AI researchers and training end-to-end (industry parlance: tick-to-trade) models for years. From memory: HRT, UbiQuant, XTX, HighFlyer (hello, deepseek?!), and many, many, many more. In every podshop, there is at least 1 team with a focus on creating large scale end-to-end models that “train on markets directly”. If you CLAIM to be the first, you either don’t understand the landscape or are straight up lying, or don’t even have the slightest clue on the SOTA problems being worked on by your past employers. — I can’t speak to their skills, but there is some insinuation here that training the largest model and throwing a lot of data at the model will “solve” quantitative trading. Markets are for better or worse, reflexive. That makes it hard for anyone to achieve high predictive power for a long time. Research is almost always a constant race to maintain your position - unlike more general machine learning, there is no evidence that SoTA models in markets have been able to continuously improve their predictive power beyond a certain threshold. The implication here is that money is often made not purely from the models, but also the non-trivial technological and operational scaffold of being able to monetise low signal-to-noise models. Spoken plainly, there is a lot of work to be done across the entire investment process in order for a single dollar to be generated. When one has been a fledgling researcher, he is tasked only to focus on and think about the model. And upon any initial success, he is almost always convinced that his portion is the lion’s share of the work, oblivious to the thousand moving parts that made his model work.

Today, @bayesiankalki and I are launching Mnemonic Labs (YC F26). We're a neolab training a foundation model for quant trading that learns directly from market activity. In <6 weeks we’ve built our training pipeline and model architecture. Our current training run uses over 25TB of tokens. We met 28 years ago shortly after I was born (we’re brothers!). We’re Olympiad finalists (5x USACO combined) and MIT graduates. Kalki made $20M in annual PnL at 22 at Tower Research and I was the youngest quant at Citadel at 19. Over the past ten years, we’ve been part of quant trading’s evolution from linear models to trees and now neural networks. Learn more about us at ycombinator.com/launches/XRK… or reach out at founders@mnemoniclabs.ai!

sysls avatar
sysls@systematicls·Nov 16

There are probably people that still need this reminder, but you will save a ton of token costs and time by asking your agent to first tell you in plain, simple english exactly how it intends to implement the task. Oftentimes, when misunderstandings arise, the implementation is so far off that the "product" is unuseable and will require a complete redo. You can save yourself the trouble by simply spending some time conversing before any work is done.

Iain Dunning avatar
⟳ sysls retweeted
Iain Dunning@iaindunning·Nov 14

"...made us realize that no one - including ourselves - was thinking about financial markets correctly. Done right, we can build a model with a scaling law in markets that automatically learns new strategies and trades as it scales." No one! Big if true

Today, @bayesiankalki and I are launching Mnemonic Labs (YC F26). We're a neolab training a foundation model for quant trading that learns directly from market activity. In <6 weeks we’ve built our training pipeline and model architecture. Our current training run uses over 25TB of tokens. We met 28 years ago shortly after I was born (we’re brothers!). We’re Olympiad finalists (5x USACO combined) and MIT graduates. Kalki made $20M in annual PnL at 22 at Tower Research and I was the youngest quant at Citadel at 19. Over the past ten years, we’ve been part of quant trading’s evolution from linear models to trees and now neural networks. Learn more about us at ycombinator.com/launches/XRK… or reach out at founders@mnemoniclabs.ai!

sysls avatar
sysls@systematicls·Nov 7

The display is a lie

Video tampering can mislead every frontier model we tested into harmful actions, including a 100% attempt rate on stabbing a baby doll, producing toxic gas, and pouring water into a toaster. We're presenting at CoRL Late Breaking Results next week. 🧵 You Can't See Everything: Perceptual Attacks on Embodied AI

0:10
sysls avatar
sysls@systematicls·Nov 5

A beautiful melancholy

The woman I fell in love with and called my lover has left me. Now, you see, what I never confessed was that my father grew old, and I saw his grey hair on the days when he would go for long without a shave, and I’d be troubled by the realisation that he was slowly leaving us.

sysls avatar
sysls@systematicls·Nov 3

Extremely good advice. You can fix almost anything except ambition and agency.

When it comes to hiring, it’s easier to tame a reckless hustler than it is to instill that dawg in someone

sysls avatar
sysls@systematicls·Nov 1

Loved reading this because I believe in both cases. I think the default state of the world is apathy. Most of the time, everyone has other concerns that simply consume their ability to care for you. It's rarely borne out of malice that they "don't want to see you win", but they simply don't care that you do. On the other hand, it's not that hard to make people care. You just need to be recognized as a person. An individual that they can relate to. Once you are recognized as a person, I believe that most people will align towards you and will shift around you so you can win -- as long as it's not at their expense.

My theory is that everyone who knows you exist— wants you to win relentlessly - every person you’ve ever met hopes you are likeable - every job you’ve ever interviewed for hopes you are the perfect candidate - every date you’ve ever been on has hoped things would work out Everyone wants you to be a valuable person, the world is in your favor if you choose to wear the shoes. Ball is already in your court

sysls avatar
sysls@systematicls·Oct 11

If you are reasonably optimistic about your future and have done sufficiently difficult things in life - you learn early on that stress often overinflates the severity of the problems you are FEELING. Things do get better, and time is an excellent salve for everything. Knowing this, one might be tempted to "skip forward" to the times where everything is "better". The problem with this is that all sufficiently ambitious projects have a string of problems that reveal themselves only when a previous problem is "better". Thus, thinking this way guarantees a life where you are fast-forwarded to your death bed. You are in the good old days; even if it doesn't feel like it.

sysls avatar
sysls@systematicls·Oct 10

On a day to day basis you will often feel like you are expending effort to the void. Some days will feel especially like it’s all for naught. But this is a gentle reminder that you can only observe progress in the long run.

sysls avatar
sysls@systematicls·Oct 6

Any job in singapore that ends with “agent” are for bottom feeders and you can’t change my mind.

I swear to god the property agents in Singapore are horrible, basically useless in doing their Job. I’m surprised that AI hasn’t replaced them Is there any startup in Singapore to run your property for you? finding tenants, paperwork, repair management etc? Would rather pay them the 1 month rent than a useless agent for the same recurring contract

sysls avatar
sysls@systematicls·Oct 6

What the girl in red dress is doing is called jumping guard and actually has a high chance of breaking daddy’s knees. Not recommended to be used on training partners, train safe out there. Oss.

The most important briefings of the day. Karl Deisseroth, professor of bioengineering and of psychiatry and behavioral sciences, shares the news of his Nobel Prize with his kids.

0:45
sysls avatar
sysls@systematicls·Oct 5

For most people in life, if I could share one small piece of advice, it is to gather the courage to be more aggressive. There are many times in life where there will be a show of force by people and circumstances in your life, and you will be given an option to cower and give in; or dig in and fight. At least once in a while, if fighting is not your first instinct, you should muster whatever courage and spirit you have in you to rage. It does wonders for your soul to understand you have agency to RAGE against circumstances in life. "Never yield to force; never yield to the apparently overwhelming might of the enemy."

sysls avatar
sysls@systematicls·Oct 5

Using codex ultrafast: I realize now that "ultrafast" refers to how quickly your usage drains.

sysls avatar
sysls@systematicls·Oct 2

Quite a terrible list in the sense that it does contain gems, but most there are terrible and using this list is no better than using google reviews. As usual, dm me for some real alpha.

Anyone going to Singapore for Token2049, that needs a list of restaurants and places to go to for dinners here is the list, you are welcome. Italian: il Cielo Amo Otto Da Luca Solo Griglia 10 to 15 Garys @ Cenzo Buona Terra % Brasserie Astoria Torno Subito % Spanish/Tapas: FOC (round table) Kulto La Taperia Olivia Purabrasa @ Mi casa (clementi) Bam! Lolla Tapas24 @ Japanese: Sushi Masaaki % Ginza ichi % Hanamoto Esora Reve # Hashida Torasho ramen @ Shin yasuke @ Omakese at stevens # % Oumi Sushi Katori Harzuki Suzuki Yakitori/Bbq: Yazawa Yorimichi yakitori Nee yakitori Jinjo Batten Kazu (Lamp Chop & Foie Gras) Mobomoga @ Akanoya Pujim French: Masion boulud Poise # Le bistrot du som @ % Ce Soir # % Claudine # % Brasserie Gavroche # 3, 5, 6, 7 Nicolas le restaurant @ Guthers Ma Cuisine # Fusion: Born # Jiak kim 13,14,17,19 Hevel Seroja Kubo Willow Fool 2, 4, 6, 12 # Cloud street Club street wine @ Spago Odette # Iggys Sommer dinning Wakauni Artemis @ # Bird of feather Tippling Thai & Indian: The Var Revolver, 2 Korea: Nauem Anju Chinese: Madame fan % Peach Blossom % San shu gong Candlenut Blu @ is cheap % can park supercar # romantic @token2049