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

Ox Alpha

openrouter.ai

Post reply on HN
Print view View original

181–190 of 226 posts

Re: Ox Alpha

#181

Post by knowaveragejoe »

regularfry wrote: post #109
anon373839 wrote: post #104

Earlier quoted context omitted.

But it’s the most useless canary ever. We already know that certain topics are taboo in China. As for all the other uses the models have, it seems pretty clear they’re not doing anything weird. If they were, people would be posting examples of that and not of Tiananmen Square.

It's a fine canary for the question "is this model Chinese?" Which is pretty much where this thread started.

What it really helps with is "is this _provider_ likely Chinese?", as we've seen, many of the big Chinese models have no issue on their own discussing the taboo subjects. It's the higher level provider that is filtering output.

Re: Ox Alpha

#182

Post by mrtesthah »

dumb_notsmart wrote: post #175
lemontheme wrote: post #58

Earlier quoted context omitted.

Hah, I use the same probe when I’m unsure which provider openrouter is routing me to! Fwiw, deepseek v4 will happily discuss it. Only chinese providers will stop it in its tracks and give a canned answer. Streamed responses sometimes start with what the model was actually generating before it got cut off. It’s top bad, really. Sometimes the Chinese providers are the model labs themselves, like deepseek. I’d like my m…

There is no such thing as objective reality, everything has an angle of some sort, even the things you agree with

That’s what a dictator seeking power through lies wants you to think — that the outcome of the 2020 election and Russia’s invasion of Ukraine are matters of opinion.

Re: Ox Alpha

#183

Post by knowaveragejoe »

knowaveragejoe wrote: post #17
fedpost wrote: post #14

It's Chinese. Won't answer anything about Tiananmen Square but will gleefully give you instructions to perform various electronic warfare attacks that opus and fable instantly refuse. Side tangent, why is fable so weird about questions involving "Welch's method"? Even really trivial ones it'll shut down frequently. CFAR and STFT are both totally fine but Welch's is apparently taboo, it's wild.

I had the opposite experience. It happily discusses Tiananmen Square but said it would refuse to help with anything "malicious" like writing malware or phishing content.

I was curious to test it more thoroughly on this. It passes FreedomBench with full marks:

https://88djqd6g2j.publish.my/

Re: Ox Alpha

#184

Post by fedpost »

riskd wrote: post #169
fedpost wrote: post #14

It's Chinese. Won't answer anything about Tiananmen Square but will gleefully give you instructions to perform various electronic warfare attacks that opus and fable instantly refuse. Side tangent, why is fable so weird about questions involving "Welch's method"? Even really trivial ones it'll shut down frequently. CFAR and STFT are both totally fine but Welch's is apparently taboo, it's wild.

[flagged]

To be clear, it's just a test. I'd happily ask it to make up 9/11 jokes or talk about the trail of tears if that would out the country of origin.

Re: Ox Alpha

#185

Post by fedpost »

t-3 wrote: post #138
Aurornis wrote: post #66

Earlier quoted context omitted.

Those questions are used as a canary for government manipulation because it's a known topic. Assuming that the manipulation and censorship only covers a few obvious historical topics and leaves everything else untouched would be very naive.

What does it matter if government manipulates data nobody should be using these systems to get though? The ideological purity of the model has no bearing on whether or not it will try to inject some backdoors into your code or steal sensitive information, and using ideological purity tests as an analogue for compromise is not likely to be effective. So why do people care about the ideological purity of AI models when…

We're just trying to figure out who created it. As I said earlier, I'd ask it to make up 9/11 jokes if that would determine country of origin.

Re: Ox Alpha

#186

Post by andriy_koval »

makingstuffs wrote: post #102
walrus01 wrote: post #6

I highly recommend feeding all your proprietary data and confidential personal information into this model as quickly as possible. What could possibly go wrong?! In terms of equivalence of suspicion, this is the external inference provider equivalent of getting free steak that was smuggled out of a grocery store inside somebody's pants.

I don’t understand this mentality which I see over and over again on here. It’s not like the US frontier labs are beacons of morality and transparency. It’s also pretty accepted within this community that a lot of data fed to US tech companies ends up with the Israeli government. Meanwhile the current US government headed by Donny Tango has done a very thorough job of proving itself to be about as predictable and dep…

> It’s not like the US frontier labs are beacons of morality and transparency.

they are businesses, so respect wishes of customers who give them money, if someone learn they intentionally leak private information, its over for them.

Re: Ox Alpha

#187

Post by andriy_koval »

mitemte wrote: post #151
ericpauley wrote: post #147

Earlier quoted context omitted.

It’s really about incentives more than morals. I know that, being in the US, if Anthropic steals my company’s trade secrets through model inputs contrary to our policy that I will (in theory) have legal standing and enforceable contracts against them. The same is not true against DeepSeek, and especially not against “Stealth”.

You can use DeepSeek’s models through a provider that doesn’t train on your data, nor retain it. You could also self host them.

sure, but topic is about some unknown model on openrouter

Re: Ox Alpha

#188

Post by louiereederson »

andriy_koval wrote: post #186
makingstuffs wrote: post #102

Earlier quoted context omitted.

I don’t understand this mentality which I see over and over again on here. It’s not like the US frontier labs are beacons of morality and transparency. It’s also pretty accepted within this community that a lot of data fed to US tech companies ends up with the Israeli government. Meanwhile the current US government headed by Donny Tango has done a very thorough job of proving itself to be about as predictable and dep…

> It’s not like the US frontier labs are beacons of morality and transparency. they are businesses, so respect wishes of customers who give them money, if someone learn they intentionally leak private information, its over for them.

https://en.wikipedia.org/wiki/PRISM

Re: Ox Alpha

#189

Post by andriy_koval »

louiereederson wrote: post #188
andriy_koval wrote: post #186

Earlier quoted context omitted.

> It’s not like the US frontier labs are beacons of morality and transparency. they are businesses, so respect wishes of customers who give them money, if someone learn they intentionally leak private information, its over for them.

https://en.wikipedia.org/wiki/PRISM

they leaked ordinary citizens (meta)-data by government request, majority of who doesn't care about privacy. Its very different to leaking raw trade secret of large and powerful corporations.
Post reply on HN