coppsilgold

7 ms·

61.
by coppsilgold 7mo ago
No one suggests the negotiated mess that exists in most standards. A single binary switch to account for hardware acceleration when it's available on both ends would have been a good decision.
62.
by coppsilgold 7mo ago
It's possible that the gains are despite the noise the coarse process introduces. After two repetitions the noise may overwhelm the advantage. The residual connections resemble the Euler method (this observation led to Neural ODE'
63.
by coppsilgold 7mo ago
It's unfortunate that WireGuard doesn't include a switch that if both sides agree the crypto in use would be AES and SHA256. Not due to FIPS compliance but performance and power savings. I never once used WireGuard on hardware tha
64.
by coppsilgold 7mo ago
Doesn't the Chinese firewall perform sophisticated filtering? Fake TCP should not be difficult to catch. I recall reading how the firewall uses proxies to initiate connections just to see whats up.
65.
by coppsilgold 7mo ago
Perhaps LLM-guided Superoptimization: < https://en.wikipedia.org/wiki/Superoptimization > I recall reading about a stochastic one years ago: < https://github.com/StanfordPL/stoke >
66.
by coppsilgold 7mo ago
> One by one your neurons are replaced by water. At which neuron are you no longer you? You have described an aspect of ageing & decline. > doesn't imply that there is no such change I made no such assertion, obviously someth
67.
by coppsilgold 7mo ago
A thought experiment: One by one your neurons are replaced by their digital counterpart as a nano-scale computer in-place, with equivalent functionality. After which neuron number are you no longer you? You remain conscious throughout the p
68.
by coppsilgold 7mo ago
Using bespoke sandboxing seems rather pointless, it will be brittle in ways you aren't going to be familiar with unless you spend time studying the bespoke method. Brittle as in it might break a workflow and you wouldn't know why,
69.
by coppsilgold 7mo ago
This just gives OpenAI that data. Perhaps you could use a local translation model to rephrase (such as TranslateGemma). If translating English to English doesn't achieve this effect then use an intermediate language, one the model is g
70.
by coppsilgold 7mo ago
For frame perfect cuts you need to re-encode. You can use lossless H264 encoding for intermediary cuts before the final one so that you don't unnecessarily degrade quality. I wonder if there is a solution which would just copy the piec
71.
by coppsilgold 7mo ago
At the time of the threats the odds were likely very skewed, as in over 99% to one side. For events where a single article could be a fulcrum, it seems like a feasible strategy to wager on the 1% and then try to manipulate the writer into c
72.
by coppsilgold 7mo ago
> End of the day, it ends up being a cat and mouse game. Cheat developers detect this and use the same resolution logic as the game client does. This is not well done. Only the server should be able to tell what the honeypot is. The poin
73.
by coppsilgold 7mo ago
> So are very good players, very bad players, players with weird hardware issues, players who just got one in a million lucky… They will all cluster in very different latent spaces. You don't automatically ban anomalies, you classif
74.
by coppsilgold 7mo ago
There is a solution to cheating, but it's not clear how hard it would be to implement. Cheaters are by definition anomalies, they operate with information regular players do not have. And when they use aimbots they have skills other pl
75.
by coppsilgold 7mo ago
Does Proton store the payment information tied to an account for the duration of a potential chargeback period or indefinitely? Whether they store such info for cryptocurrency payments as well (no chargeback risk) would be telling.
76.
by coppsilgold 8mo ago
Some of the best current voice tokenizers achieve ~12 Hz, that's many more tokens than a regular LLM would use for ultimately the same content.
77.
by coppsilgold 8mo ago
The author seems obsessed with RaptorQ[1], this is not a good place for it. RS over GF256 is more than adequate. Or just plain LDPC. [1] < https://www.jeffreyemanuel.com/writing/raptorq >
78.
by coppsilgold 8mo ago
Containers, virtual machines, jails. Containers have jail runtimes available. See for example gVisor/runsc.
79.
by coppsilgold 8mo ago
At the moment LLM's tend to work well when you constrain them, and you can craft the constraints with the help of the same LLM in a different session. Then you can verify if the outputted code obeys the constraints in yet another sessi
80.
by coppsilgold 8mo ago
The difference between gVisor and a microVM isn't very large. gVisor can even use KVM. What gVisor doesn't have is the big Linux kernel, it attempts to roll a subset of it on its own in Go. And while doing so it allows for more co
81.
by coppsilgold 8mo ago
If you are on Linux you can 'download' some RAM. Enable zram, configure sysctl variables to make good use of it. Note that it won't help you if your workload makes use of all your RAM at once. If you have a bunch of stuff run
82.
by coppsilgold 8mo ago
Private keys from secure elements leak all the time. There will be a flawed implementation that someone exploits, an insider will smuggle a key out etc. This is why true zero-knowledge systems for this sort of thing aren't practical an
83.
by coppsilgold 8mo ago
How feasible would it be to integrate a neural video codec into the SoC/GPU silicon? There would be model size constraints and what quality they can achieve under those constraints. Would be interesting if it didn't make sense to
84.
by coppsilgold 8mo ago
You don't even need active measures. If a publisher is serious about tracing traitors there are algorithms for that (which are used by streamers to trace pirates). It's called "Traitor Tracing" in the literature. The ide
85.
by coppsilgold 8mo ago
Performance like that may open the door to the strategy of brutefocing solutions to problems for which you have a verifier (problems such as decompilation).
86.
by coppsilgold 8mo ago
You can buy Psyllium husks by the pound as well.
87.
by coppsilgold 8mo ago
> There has never been a perfect CSPRNG. What is a perfect CSPRNG?
88.
by coppsilgold 8mo ago
Cryptographic operations when done correctly result in full chaos within the discrete domain (the so-called avalanche effect). Any bias of any kind gives rise to a distinguisher and the primitive is regarded as broken. One way to imagine wh
89.
by coppsilgold 8mo ago
Does there exist a single MCMC example that performs poorer when fed by a CSPRNG (with any fixed seed, including all zeroes; no state reuse within the simulation) as opposed to any other RNG source?
90.
by coppsilgold 8mo ago
You can roll a script to do this, something that would consume a mic from Pipewire when triggered and then push results to clipboard. With a Parakeet ONNX model in between. I had cause to do the the opposite: Hotkey -> clipboard TTS

More ›