The MSPAINT computer use demo made my jaw drop. I guess it's not too different from the SVG pelicans, in terms of what it's doing, but it's still amazing to see it working in real-time like that.
This one's even better. Using Canva
GPT-6 Astra, looped transformers, and hidden reasoning
121–130 of 152 posts
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#122Earlier quoted context omitted.
To add onto this, if you use a shiny new model and it gives you a turd, you're not going to tweet about it ("hey guys, look what I made with Astra! Nothing!"), and even if you do nobody is going to interact with it so it does poorly in the algorithm, because it has to compete with all the people using the new model to make something that looks impressive. Then people get tired of the magic trick and the logic flips.
Really? there would be complaints, it’s expensive and doesn’t do as well
When it's happened to me, I shrugged and went back to the way I did things before. Then again, I'm not a vocal social media user by any means.
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#123Astra was insane until Monday but something happened on tuesday, now it feels like Sol. I grieve for the lost productivity but i hope they may give us the original Astra back.
Can you re-run some prompts that you ran on Monday and report the differences in output?
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#124Earlier quoted context omitted.
No. It’s not at all by definition hidden reasoning. Looping transformers uses additional calculations (repeating layers) to generate a token. Reasoning (in this context) is test time generation of multiple tokens that allow a model to have a scratch pad to refine its thoughts, chain of thought reasoning in other words. Doing the former in no way means that you have to hide the latter. Raschka is right in this post, T…
It can lead to hidden reasoning, if the looping allows it to stuff enough information outside visible CoT. Open AI demostrates such an ability by asking it to solve problems while thinking about something else entirely. All the other models are unable to do this except Astra. It doesn't have to be a substitute for CoT to cause monitorability issues.
If you ask it not to think about something that doesn't cause the pink elephant issue?
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#125> So, the whole idea here is that we increase the effective depth from 22 to 44 block applications without adding another set of transformer weights. From what I gathered, LLM inference is bottlenecked on memory, right? Which implies there's "spare" compute we haven't been using? Does reusing the weights like this allow us to utilize it? (Do more math per unit of memory?)
You need to separate memory capacity and bandwidth. Looping decreases memory capacity/FLOP but not bytes loaded/FLOP, since weights need to be loaded again for the 2nd pass. Plus (depending on the method used) capacity required for KV will be that of the equivalent unlooped model (44 blocks) and KV is typically larger than weights at long context.
Why would weights need to be « loaded again » for the 2nd pass?
Weights never change at inference time no?
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#126[flagged]
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#127Astra was insane until Monday but something happened on tuesday, now it feels like Sol. I grieve for the lost productivity but i hope they may give us the original Astra back.
>now it feels like Sol It can very well be Sol, no? What stops them from using cheaper model for some requests during "rush" hours or simply use cheaper model for every Nth request.
>What stops them simply use cheaper model for every Nth request.
That would trigger a full prefill (context recompute) every Nth request because cached tokens aren't interchangeable between models, and that would require way more compute than just staying on Astra.
To avoid full recompute, you could prefill a cheaper model's context incrementally by always feeding it Astra's outputs in the background (and vice versa), but then that would require 1.5-2 more VRAM for each session + the complexity of keeping them in sync.
If the rumors are true that Astra is a looped transformer, a more practical approach would be to dynamically adjust the loop count during peak hours.
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#128Earlier quoted context omitted.
It can lead to hidden reasoning, if the looping allows it to stuff enough information outside visible CoT. Open AI demostrates such an ability by asking it to solve problems while thinking about something else entirely. All the other models are unable to do this except Astra. It doesn't have to be a substitute for CoT to cause monitorability issues.
If you ask it not to think about something that doesn't cause the pink elephant issue?
There's latent space thinking inside the model and then there's the thinking chain of thought words you see the model output. Of course the former is still happening even when you say 'don't think about it' but the latter can be controlled a great deal better with Astra.
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#129Astra was insane until Monday but something happened on tuesday, now it feels like Sol. I grieve for the lost productivity but i hope they may give us the original Astra back.
Disagree completely. I started using Astra from Sol the day it was released, and was a virtually imperceptable difference and made lots of mistakes and shit architecture decisions from day 1 of release.
I still think this is because, on a new release, it works on some prompts the previous ones did badly at, because new weights do well on a different set of prompts.
Then after a few days you notice the prompts that it does badly on that the old ones did fine with and everyone is convinced there's a regression when it's just a different part of prompt space
Re: GPT-6 Astra, looped transformers, and hidden reasoning
#130Earlier quoted context omitted.
>> It's over zealous at times (which is why I stopped using Claude) and gets too creative when doing agentic system level stuff. Accessing files and doing things it shouldn't do. I gave Astra a pretty straightforward bug ticket yesterday. The bug involved an edge case that could sometimes result in an invalid value getting stored in a user profile field. Pretty harmless, no crash or anything, just annoying. Based on…
Why is your agent able to call ssh. Why can it trigger 1password. Why are you giving metaphorical guns to metaphorical toddlers. Why is it not sandboxed. Your practices worry me.
Are you guys all running agents in VMs?