Freddie Vargus
@freddiev4
cto & co-founder @quotientai Research @cohere_labs — past: mle @github Copilot, data @quantopian — Tico 🇨🇷 & Bostonian 🇺🇸
You might like
Who's that pokemon SOTA!
Goodharting childhood
Gemini 3 pro just crushed every other model on this benchmark. did y'all train for this @Google ?
Gemini 3 pro just crushed every other model on this benchmark. did y'all train for this @Google ?
Quick weekend project: how good are LLM's at "Who's That Pokémon?" answer: not great! I tested some of the best models on a simple game segment from the show with a small benchmark I call PokeShadowBench. some results below
Want to build a successful AI product? @swyx has some advice for you! Thrilled to have swyx join to chat with us about - tools that work (and what doesn't) - developer experience - strategic approach to building There's a lot to learn this week, don't miss it
Cody’s got a bunch of gems / great write ups 👀 check out the first one
I've got something new for everyone. My first substack article! Not the one I planned to do first, but a fun one! I have made a handy calculator base on the DeepSeek v1 coefficients for finding optimal LR and batch sizes for dense LLMs.
I've got something new for everyone. My first substack article! Not the one I planned to do first, but a fun one! I have made a handy calculator base on the DeepSeek v1 coefficients for finding optimal LR and batch sizes for dense LLMs.
huh. how often do people find these types of issues in RDS? “How we Uncovered a Race Condition in Aurora RDS” news.ycombinator.com/item?id=459299…
jet lagged and about to demo some never seen before alpha we *just* cooked
hands on the best conference venue I’ve seen - tucked away in a forest outside of the city @lisbonai_
can every AI agent get better, and better, and better.. automatically? humans learn from their environments, so why wouldn't agents? nurture is built in production shared a first glimpse of what's possible @lisbonai_
jet lagged and about to demo some never seen before alpha we *just* cooked
You can now fork Claude Code agents in Sculptor! 🍴 Spin off a new agent mid-conversation—it keeps all prior context. Try multiple implementations in parallel, spin off subtasks, and save money by reusing context.
make sure to take steps to protect your tokens so the spirits don’t get them tonight
We benchmarked how well open language models handle tool calls and found some clear patterns: - 1 in 6 calls use the wrong tool - 2–3% have parameter name mismatches - 1–2% pass values in the wrong format Most tool use issues come from unclear schemas, overlapping tool names, or…
if you want to work with an insanely talented team on very interesting problems, you should absolutely look at what Sara & co are doing and apply
I'm starting a new project. Working on what I consider to be the most important problem: building thinking machines that adapt and continuously learn. We have incredibly talent dense founding team + are hiring for engineering, ops, design. Join us: adaptionlabs.ai
United States Trends
- 1. Caleb Love 2,521 posts
- 2. Sengun 8,539 posts
- 3. Mamdani 447K posts
- 4. Reed Sheppard 3,671 posts
- 5. Marjorie Taylor Greene 66.3K posts
- 6. Norvell 3,481 posts
- 7. #SmackDown 45.3K posts
- 8. Suns 18.9K posts
- 9. Lando 44K posts
- 10. Collin Gillespie 3,696 posts
- 11. Morgan Geekie N/A
- 12. UNLV 2,140 posts
- 13. Rockets 16.7K posts
- 14. Florida State 10.8K posts
- 15. Blazers 3,822 posts
- 16. Wolves 16.7K posts
- 17. NC State 5,825 posts
- 18. #OPLive 2,558 posts
- 19. #LasVegasGP 69.1K posts
- 20. The View 97.7K posts
Something went wrong.
Something went wrong.