Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Can you reverse engineer our neural network?

blog.janestreet.com

Post reply on HN
Print view View original

71–80 of 216 posts

Re: Can you reverse engineer our neural network?

#71

Post by wat10000 »

logicchains wrote: post #43
dgacmu wrote: post #37

Earlier quoted context omitted.

We do have a lot of policy levers that are tilted in favor of making money in finance, though, and we could change those levers. - The carried interest loophole - We could add small transaction taxes - We could raise capital gains taxes - We could be a little more focused on enforcing antitrust - We could raise higher end marginal tax rates to reduce the relative attractiveness of off-scale payrate jobs - We could pr…

>long-term societal importance without bringing in any form of centralized resource allocation. The onus is on the biomedicine industry to demonstrate it's capable of producing anything of societal importance because so far it's largely failed to deliver. There's nothing noble or scientific about throwing good money after bad into an industry that's continuously failed to deliver.

Antibiotics? Vaccines?

Re: Can you reverse engineer our neural network?

#72

Post by clouedoc »

I'm really curious what were the magic words.

> Alex had actually tried to brute force the hash earlier, but had downloaded a list of the top 10,000 most popular words to do it, which turned out not to be big enough to find it. Once he had a big enough word list, he got the answer.

They don't reveal the answer.

Re: Can you reverse engineer our neural network?

#73

Post by airza »

alehlopeh wrote: post #68
wittyusername wrote: post #5

Earlier quoted context omitted.

All I think when I see this is "this intelligence wasted on finance and ads." Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? Not everything has to be perfectly efficient but it just saddens me to see all these great minds doing what, adversarially harvesting margin from the works of others?

This kind of thinking denies the humans in question any agency over their own lives. You’re essentially asking for a word government and a planned economy. Does it suck that more resources aren’t funneled into cancer research and other noble pursuits? Yes. But planned economies haven’t cured cancer, and are not likely to, despite denying their people the agency to live their lives as they choose.

I don’t think you have to go from here to planned economy straight away. There are capital gains taxes between the current level and 100% which might produce better outcomes.

Re: Can you reverse engineer our neural network?

#74

Post by vecter »

throwaw12 wrote: post #20
JasonADrury wrote: post #15

Earlier quoted context omitted.

High frequency quants almost certainly don't directly provide a massive social good, but they're one of the many facilitators enabling the smooth functioning of markets that do. But FWIW, the comment I was replying didn't seem to be specifically critical of high frequency quants. Dismissing the entire field as something that doesn't contribute to society is beyond absurd.

> facilitators enabling the smooth functioning of markets that do Do you think without HFT markets stop functioning properly? What if they are one of the contributors of highly inflated valuations?

HFT doesn't inflate or deflate any valuations. They operate at the market microstructure level and provide liquidity. HFT firms have no impact on the long-term value of assets.

Re: Can you reverse engineer our neural network?

#75

Post by pixl97 »

relma2 wrote: post #54
wittyusername wrote: post #5

Earlier quoted context omitted.

All I think when I see this is "this intelligence wasted on finance and ads." Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? Not everything has to be perfectly efficient but it just saddens me to see all these great minds doing what, adversarially harvesting margin from the works of others?

"But I don't want to cure cancer. I want to turn people into deepfakes."

The thing is the same hardware does both, Deepmind should be the best example of a company that's developed both.

Also there is no such thing as 'cancer' it is '_____ cancer' which seems to be about one of a million different things. So technically you have to cure cancer a million times. Expanding our compute ability and smart algorithms to solve the medical issues are a far more scalable option than having humans work on it alone.

Re: Can you reverse engineer our neural network?

#76

Post by ericd »

state_less wrote: post #58
wittyusername wrote: post #5

Earlier quoted context omitted.

All I think when I see this is "this intelligence wasted on finance and ads." Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? Not everything has to be perfectly efficient but it just saddens me to see all these great minds doing what, adversarially harvesting margin from the works of others?

> Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? We already have very efficient crop harvesting and Eli Lilly is nearly a $1 Trillion dollar company. Interestingly, the new medicine is designed to keep us from eating so many cheap calories (new weight loss drugs). > Not everything has to be perfectly efficient but it just saddens me to see all these great…

It’s an important function, but the guys making these bets frequently pull down $10-$100M per year, each. That’s a huge toll to extract from the productive economy for playing this game.

And then there are the guys managing things like pensions, skimming a percentage every year, just because they happen to be locked into that position, meanwhile underperforming a basket of index funds. Just happily eating away at the retirement savings of thousands-millions.

Re: Can you reverse engineer our neural network?

#77

Post by gormen »

hal9000xbot wrote: post #30
dang wrote: post #175

[stub for offtopicness]

[flagged]

I approached the puzzle using an A11‑style reasoning architecture, which focuses on compressing the hypothesis space rather than decoding every neuron. Instead of “understanding the network”, the task reduces through successive narrowing: model → program → round‑based function → MD5 → dictionary search for the target hash. The key steps were:

Input → the integer weights and repeated ReLU blocks indicate a hand‑designed deterministic program rather than a trained model.

Weighting → the only meaningful output is the 16‑byte vector right before the final equality check.

Anchor → the layer‑width pattern shows a strict 32‑round repetition, a strong structural invariant.

Balancing → 32 identical rounds + a 128‑bit output narrow the function family to MD5‑style hashing.

Rollback → alternative explanations break more assumptions than they preserve.

Verification → feeding inputs and comparing the penultimate activations confirms they match MD5 exactly.

Compression → once the network becomes “MD5(input) == target_hash”, the remaining task is a constrained dictionary search (two lowercase English words).

The puzzle becomes solvable not by interpreting 2500 layers, but by repeatedly shrinking the search space until only one viable function family remains. In this sense, the architecture effectively closes the interpretability problem: instead of trying to understand 2500 layers, it collapses the entire network to a single possible function, removing the need for mechanistic analysis altogether.

Re: Can you reverse engineer our neural network?

#78

Post by vonneumannstan »

state_less wrote: post #58
wittyusername wrote: post #5

Earlier quoted context omitted.

All I think when I see this is "this intelligence wasted on finance and ads." Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? Not everything has to be perfectly efficient but it just saddens me to see all these great minds doing what, adversarially harvesting margin from the works of others?

> Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? We already have very efficient crop harvesting and Eli Lilly is nearly a $1 Trillion dollar company. Interestingly, the new medicine is designed to keep us from eating so many cheap calories (new weight loss drugs). > Not everything has to be perfectly efficient but it just saddens me to see all these great…

I think the comment was a roundabout way of saying this is a clear market failure. There are more societally important things these people could be doing instead of shaving another ms off a transaction or finding minuscule option pricing inefficiencies. That the market is not correctly remunerating those options is the failure.

> For example, the tariff tantrums caused by trump proposing 100%+ china tariffs where he crashed the markets last spring, leading to a moderation in policy.

"Akshually traders are good bcuz they crash the market when the president does insane things" is not the own you think it is.

Re: Can you reverse engineer our neural network?

#79

Post by pixl97 »

logicchains wrote: post #47
bluerooibos wrote: post #31

Earlier quoted context omitted.

I'm sure people (society) wants cheap food, free universal healthcare, free public transport, so why don't we have these things? Under capitalism each of these individual systems needs to turn a profit to be deemed worthwhile instead of treating the system as a whole and taking into account the economic externalities and benefits to the entire system.

>free universal healthcare, free public transport There's no such thing as free things, there's just some people paying for other people's things, and surprise surprise some people don't want to use their hard earned money to pay for other people's transport and healthcare.

And yet too much zero sum thinking leads to a crabs in a bucket mentality were the greedy get less by being greedy instead of having an educated productive society around them.

Re: Can you reverse engineer our neural network?

#80

Post by SkyBelow »

autonomousErwin wrote: post #36
wittyusername wrote: post #5

Earlier quoted context omitted.

All I think when I see this is "this intelligence wasted on finance and ads." Can you imagine human potential if it was somehow applied to crop harvesting efficiency, new medicines, etc? Not everything has to be perfectly efficient but it just saddens me to see all these great minds doing what, adversarially harvesting margin from the works of others?

I think this is the wrong way to think about it. In this case the "intelligent people who are wasted on finance and ads" are drawn to high-status, low-risk, well-paid jobs, with interesting problems to solve. If you want to solve meaningful problems you need a different kind of intelligence; you need to be open to risk, have a lot of naivety, not status orientated, and a rare ability to see the forest among the trees…

>not status orientated

Should we view those who chase status as a bad thing, or look to those who assign status that is then chased? If the average person cares more about who won last night's big game than some work done to improve medication, should we really have anything to say about those who decide to optimize their lives by what society actually rewards?

I noticed this way back in grade school. Good grades were, if anything, a net negative prestige, while sports were a positive prestige. It made me wonder what the school was actually optimizing for, because the day to day rewards weren't being given to the studious. (The actual reward function was more complicated, such as good grades being a boost if one was already a sports star, but these were exceptions to the norm.)

Post reply on HN