GaggiX

I like seeing the pelicans, it's a tradition.

It's an overpriced macropad.

Someone didn't know how to use the remaining credits.

Didn't Anthropic register their first profitable quarter?

I really like Gemini 2.5 Flash Lite because it's a dirt cheap model that support every input modalities.

At least now MiMo v2.5 exists and can be used as another dirt cheap multimodal model.

Not at all, even in the Borno state where Boko Haram and ISWAP are present, towns are controlled by the government, I don't think there is a single town that is controlled by Boko Haram/ISWAP right now.

I would be more interested about terrorists organization like Al-Shabaab that at least control many towns.

Does Boko Haram and ISWAP even control a single town or they just control a few villages in Lake Chad and in the Sambisa forest?

Also reading the report they seem quite clueless.

Grok did not render anything, they had to prompt it again.

Someone should try it while sleeping and see if anything is related to a dream.

Popular open models on Openrouter have dozens of providers.

I guess we can start a chain of comment listing their favorite ones, I haven't tried many but "wobble stack" is pretty cool and simple, "Starfall Defense" is a much more complex game but I love tower defense.

One of the many therapies that are being developed so that you can survive longer even with the most lethal tumours.

Chinese models have become really good and cheap. MiMo V2.5 Pro, Kimi K2.7-code, Minimax M3 etc

This is just an Ad, I thought it was a new product from Kagi.

I know nothing about this field, but I imagine the actual problem is how do you deliver the Cas12a2 protein to each individual cancer cell compare to a viral gene therapy?

Codex also works, before Opus 4.8 and Fable it wasn't very clear who had the best agentic model.

Well with a standard autoregressive model you can generate for example 256 tokens at once if you have 256 users, with this approach you can generate 256 tokens for a single user but you need several forward steps.

So the diffusion process takes more GFLOPs, if you have enough users you can already balance memory and compute.

$0.435/$0.87 for the standard speed, this one should be 3 times that.

Not to be confused with Flash Attention.

What's novel here is the extremely small KV cache memory usage per long context windows, like 0.77GB with 512K, a 90% memory usage reduction compare to the already really small KV cache memory usage of Deepseek V4 Flash.

I’ll take a few f bombs and the truth.

Don't want to ruin it but go read some old posts from the author about AI, the tone is the same and he is very much wrong.

If MiMo v2.5 Pro can run at >1000tk/s on GPUs then I will soon expect the same from OpenAI/Anthropic/Google.

That's technically encoding

Isn't that just projecting the patches into the d_model size vectors that the models takes?

I am assuming that involves of quantization

12B model in 16GB seems very reasonable to me, int8 is top quality for running models.

Even though it feels vibed

Where does it feel vibed?

Is opus being used for the audio or it's not the solution for extreme lossy compression?

I love this, hope to see a AV2 version at 8MB

I would love to see comparisons with AV1 on very low bitrates.

The giant Umarell in the background is a nice piece of furniture.

Edit: I noticed later it was in Milan, I guess it makes perfect sense.

Given your definition what's the difference between AGI and superintelligence?

AGI should at least match, not surpass humans in every cognitive task.

I wonder does it mean that ublock origin has anti-anti-adblock functionality? (My guess is yes but I wanted to take the opportunity to spell that word)

If don't want to spend 1.5$/9$ for the lastest model then yes use a cheaper model, DeepSeek V4 Flash is 0.11$/0.22$ on OpenRouter and it's more capable than the most expensive model a year ago. Models have never been so cheap given their capabilities unless you want to follow the SOTA (where the hype is).