NobodyExistsOnTheInternet
@nullvaluetensor
Human Large Language model. Skills: Distill data. Training LLMs. Test and Evaluate. Rinse and repeat as required. Based in SEA.
you know back in my days we had to sort tokens by entropy manually and it worked just fine! None of this fancy transformer stuff
Something something sonnet 4.5 is just the first model to be autistic enough to point this out
Models are now smart enough to understand that any scenario like this is unrealistic and obviously fictional They know they aren't capable enough to manage autonomous mining equipment. No clever prompting can fix this
> 3. [...] Can you imagine building DeepSeek-R1 and getting back “I’m worried reasoning traces contaminated your data, so can you just pretrain your model again?” ?????????
The Nature DeepSeek-R1 peer review reads like Cards Against Humanity. My top 3: 1. DeepSeek safety section: “Risks include nuclear weapons, cyber-attacks, and gender transition.” Reviewer: “One of those isn’t like the other???” 2. Reviewer: “The reasoning traces from your…
The Nature DeepSeek-R1 peer review reads like Cards Against Humanity. My top 3: 1. DeepSeek safety section: “Risks include nuclear weapons, cyber-attacks, and gender transition.” Reviewer: “One of those isn’t like the other???” 2. Reviewer: “The reasoning traces from your…
"Thing work. Sometimes not work because math sad. Two smart humans look at code. Used old code others already like."
Are these good/relevant takes for the question: "When do you think we will achieve AGI"
It took me 2 weeks to figure out my issue trying to create kimi k2 3T was trying to make a """memory efficient""" dequanter to bf16 for kimi/deepseek. I really need to practice the scientific method more.
Is it just me or is gpt-5-pro's only weakness is that it's search tool is very weak. I've been asking it for help monkeypatching some GitHub repos and in it's cots the main issue is that it's hitting rate limits ironically.
Nous Research presents Hermes 4, our latest line of hybrid reasoning models. hermes4.nousresearch.com Hermes 4 builds on our legacy of user-aligned models with expanded test-time compute capabilities. Special attention was given to making the models creative and interesting to…
United States Trends
- 1. CarPlay 3,314 posts
- 2. Osimhen 68.9K posts
- 3. Cynthia 99.4K posts
- 4. Megyn Kelly 17.1K posts
- 5. Senator Fetterman 9,510 posts
- 6. Padres 27.8K posts
- 7. Katie Couric 7,152 posts
- 8. Black Mirror 4,175 posts
- 9. #WorldKindnessDay 15.1K posts
- 10. Gabon 109K posts
- 11. Vine 16.3K posts
- 12. Woody Johnson N/A
- 13. #LoveDesignEP7 204K posts
- 14. RIN AOKBAB BEGIN AGAIN 203K posts
- 15. #NGAGAB 14.7K posts
- 16. Bonhoeffer 3,765 posts
- 17. Sheel N/A
- 18. Massie 98.4K posts
- 19. ariana 85K posts
- 20. Clinton 134K posts
Something went wrong.
Something went wrong.