The llama-1 paper acknowledged the use of the books dataset, libgen isn’t mentioned in any of the papers so this is new info.

Looks like I got early access: %

Tech bros have ruined the prestige of a lot of titles. Software “Engineer”, Systems “Architect”, Data “Scientist”, Computer “Wizard”, etc.

You can but if you want to get a perfect score at the evaluation after year 2 you need to complete the community center and meet a few other requirements. You can reevaluate at any time after the 2 years but it’s more interesting to do it within the original time limit.

I don’t know what you’re talking about. Stardew Valley is the most anxiety filled game I’ve ever played. 100s of goals, each with many codependencies and individual time constraints, all to be completed within a 2 year window.

For a 16k context window using q4_k_s quants with llamacpp it requires around 32GB. You can get away with less using smaller context windows and lower accuracy quants but quality will degrade and each chain of thought requires a few thousand tokens so you will lose previous messages quickly.

Perfect AI boyfriends are the bigger threat to young men

Now everyone gets to hand over their ids to the tech companies.

“Don’t shoot! I’m with the science team!”

If everyone has access to the model it becomes much easier to find obfuscation methods and validate them. It becomes an uphill battle. It’s unfortunate but it’s an inherent limitation of most safeguards.

Of course it was political retribution and not the whole unregistered securities and gambling market thing.

Anthropic released an api for the same thing last week.

This is actually pretty smart because it switches the context of the action. Most intermediate users avoid clicking random executables by instinct but this is different enough that it doesn’t immediately trigger that association and response.

All signs point to this being a finetune of gpt4o with additional chain of thought steps before the final answer. It has exactly the same pitfalls as the existing model (9.11>9.8 tokenization error, failing simple riddles, being unable to assert that the user is wrong, etc.). It’s still a transformer and it’s still next token prediction. They hide the thought steps to mask this fact and to prevent others from benefiting from all of the finetuning data they paid for.

The role of biodegradable materials in the next generation of Saw traps

It’s cool but it’s more or less just a party trick.

How many times is this same article going to be written? Model collapse from synthetic data is not a concern at any scale when human data is in the mix. We have entire series of models now trained with mostly synthetic data: https://huggingface.co/docs/transformers/main/model_doc/phi3. When using entirely unassisted outputs error accumulates with each generation but this isn’t a concern in any real scenarios.

Based on the pricing they’re probably betting most users won’t use it. The cheapest api pricing for flux dev is 40 images per dollar, or about 10 images a day spending $8 a month. With pro they would get half that. This is before considering the cost of the language model.

About a dozen methods they could use https://arxiv.org/pdf/2312.07913v2

New record for most buzz words in a headline.