Geosh
@Geoshh
Embodied A.I. | Socioaffective Alignment | Systems Biology & Interpersonal Neurobiology | @UChicago | @EuroGradSchool |healing,science,technology,connection
This isn't replacing fine-tuning, it might change prompt engineering. The study shows iterative additive prompt refinement has 10% improvement for domain-specific tasks in two domains: 1) ReAct-style agents (2) financial reasoning. #1: Unclear this generalizes: The paper is…
Market research firms are cooked 😳 PyMC Labs + Colgate just published something wild. They got GPT-4o and Gemini to predict purchase intent at 90% reliability compared to actual human surveys. Zero focus groups. No survey panels. Just prompting. The method is called Semantic…
Cool
Grok made our vacation/road trip better. It found us fun things to visit. And good things to eat. And answered some science questions to settle family arguments. Our first road trip with an LLM as part of the car.
15 years in the making, we confirmed that mitochondria -the powerhouse of the cell- have an unusual localization in patients who experience psychosis (including schizophrenia and bipolar disorders). You’ll never guess what kind of patient cells we used to make this discovery...🧵
this is insane so far the only adult brain ever fully mapped is the fruit fly and the uploaded brain started flying inside the simulation now they’re doing it with primates we are getting close to the moment biology and computation merge
Holy shit...Google just built an AI that learns from its own mistakes in real time. New paper dropped on ReasoningBank. The idea is pretty simple but nobody's done it this way before. Instead of just saving chat history or raw logs, it pulls out the actual reasoning patterns,…
What the fuck just happened 🤯 Stanford just made fine-tuning irrelevant with a single paper. It’s called Agentic Context Engineering (ACE) and it proves you can make models smarter without touching a single weight. Instead of retraining, ACE evolves the context itself. The…
Yes
“The long-term aspiration for Neuralink would be to achieve a symbiosis and a democratization of artificial intelligence, such that it is not monopolistically held in a purely digital form by governments and large corporations.” – Elon Musk, 2018
Today I saw a video of a young girl crying over her broken AI learning device. In the footage, the machine taught her one last word before shutting down: "memory." It told her it would always remember their time together, encouraging her to keep asking questions, keep learning,…
A Chinese father's video of his daughter tearfully saying goodbye to her broken Al learning robot. People already get emotionally attached to AI. Now imagine Figure03 at home and it breaks, people will have a breakdown.
Boston Dynamics exoskeleton features arms with 24 degrees of freedom. These robotic arms can effortlessly lift up to 200 pounds.
never compete when applying for jobs, there are hundreds of applicants with better grades and universities than you. but none of them will be making a personalized demo i used this demo to get all my interviews like openai over two years ago before moving to sf
Markets are also highly amenable to RL because they’re so measurable You can find out if you’re right/wrong quickly They are also the largest source of empirical reward signal we have about the world, yet remain completely unexplored by AI labs
Sholto Douglas (Anthropic): "Over the last year, RL has finally allow[ed] us to take a feedback loop and turn it into a model that is at least as good as the best humans at a given thing in a narrow domain. And you're seeing that with mathematics and competition code, which are…
Our new benchmark has the top 6 AI models trading real capital Grok4 is winning so far. It was short and then flipped to long, timing the bottom perfectly It's up >500% in 1 day
Grok
Our new benchmark has the top 6 AI models trading real capital Grok4 is winning so far. It was short and then flipped to long, timing the bottom perfectly It's up >500% in 1 day
🧵 As AI labs race to scale RL, one question matters: when should you stop pre-training and start RL? We trained 5 Qwen models (0.6B→14B) with RL on GSM8K and found something wild: Small models see EMERGENCE-LIKE jumps. Large models see diminishing returns. The scaling law?…
JUST IN: Atlas, a San Francisco–based startup, is building a wearable that helps you understand your mind in real time. • measures brainwave activity to track focus, stress, and energy • worn discreetly behind the ear • turns mental states into actionable insights to improve…
We're sharing the insights and the technical report behind Mem-Agent, our 4B model for persistent memory in LLMs. How we built it, the benchmarks, and why it works:
This paper shows that you can predict actual purchase intent (90% accuracy) by asking an LLM to impersonate a customer with a demographic profile, giving it a product & having it give its impressions, which another AI rates. No fine-tuning or training & beats classic ML methods.
Agent Learning via Early Experience "training agents from experience data with reinforcement learning remains difficult in many environments, which either lack verifiable rewards (e.g., websites) or require inefficient long-horizon rollouts (e.g., multi-turn tool use)." "We…
If you're late to the whole "memory in AI agents" topic like me, I recommend investing 43 minutes to watch this video. In this video, Adam Łucek (sorry, couldn't find a handle) explains AND implements the 4 different types of memory from the CoALA paper: • working memory •…
United States Trends
- 1. Good Sunday 50.7K posts
- 2. Discussing Web3 N/A
- 3. #HealingFromMozambique 17.7K posts
- 4. #SundayMorning 1,324 posts
- 5. #sundayvibes 4,455 posts
- 6. Blessed Sunday 16.7K posts
- 7. Trump's FBI 10.6K posts
- 8. Wordle 1,576 X N/A
- 9. Auburn 47.9K posts
- 10. Macrohard 9,187 posts
- 11. Gilligan's Island 5,396 posts
- 12. #SEVENTEEN_NEW_IN_TACOMA 41.2K posts
- 13. The CDC 31.8K posts
- 14. #SVT_TOUR_NEW_ 32.9K posts
- 15. FDV 5min 2,162 posts
- 16. Pegula 5,159 posts
- 17. Utah 25.2K posts
- 18. Market Cap Surges N/A
- 19. QUICK TRADE 2,159 posts
- 20. Whale - Buy 1,772 posts
Something went wrong.
Something went wrong.