AssistedEvolution
@AssistedEvolve
Thinking outside of the brane that contains the box
You might like
Prominent Quantum Computing Use Evaporates A group of physicists argues that one of the most-often used examples for the use of quantum computers --- solving the travelling-salesman problem --- is a futile attempt. They reviewed two decades of failed attempts to force the…
We can now map the entire dendritic structure of neurons in a brain. It's called dendritome mapping. And it's been used to characterize dopaminergic neurons in the part of the brain most affected in schizophrenia. Published in @NatureNeuro Oct 31 2025
Pinghu Lab unveiled GaN MPW (Multi Project Wafer) service allowing each wafer to test multiple designs in 25mm2 units. Paired w/ GaN PDK to create & tape out chips that they can test out. 650V MPW to launch this yr & 100V MPW by 2026Q1. Cell library includes GaN HEMT, resistors,…
The dysregulated immune system in #LongCovid with #MECFS. with sex-specific changes, increased inflammation (as seen by cells, chemokines and cytokines), and disrupted hormone levels cell.com/cell-reports-m…
Bayesian methods don't work. People often ask what's wrong with Bayesian methods, but the issues have been well-documented for years. Here are some key points: 1) Bayesian confidence values lack theoretical basis. 2) They rely on unjustified priors. 3) They produce posterior…
Shenzhen alone has so many amazing robotics companies.
It’s actually pretty insane how many humanoid companies there are. Feels like a new one pops up every week
Huge AI news from Google Google Research just unveiled a bold new ML paradigm that views a model as a stack of nested problems so it can keep learning new skills forever without forgetting the old ones, a huge leap toward AI that actually evolves like a brain. 👀 On tests of…
Google introduces Nested Learning, "a new ML paradigm for continual learning": "Nested Learning... represents a step forward in our understanding of deep learning. By treating architecture and optimization as a single, coherent system of nested optimization problems, we unlock a…
Introducing Nested Learning: A new ML paradigm for continual learning that views models as nested optimization problems to enhance long context processing. Our proof-of-concept model, Hope, shows improved performance in language modeling. Learn more: goo.gle/47LJrzI…
This is a really excellent article by AYZ about ASICs. Apart from the excellent content of the article itself, I found myself wondering about something. It is said that Google directly designed the ASIC die for TPU v7e. This is quite different from the rumor circulating that…
Marvell (MRVL US) vs. Broadcom (AVGO US) vs. Alchip (3661 TT) vs. GUC (3443 TT) – A Brief Update on ASIC Plays open.substack.com/pub/globaltech…
We're getting closer, we're making progress👀 Physicists have found strong evidence of unconventional superconductivity in a special form of graphene ('magic angle' twisted trilayer graphene), a discovery that could pave the way toward room temperature superconductors.
Physicists observed key evidence of unconventional superconductivity in a special form of graphene. The findings may guide the design of superconductors that work at room temperature, “which is sort of the Holy Grail of the entire field,” Pablo Jarillo-Herrero says.…
This is one of the most exciting agentic AI results I have seen! An AI agent (through several rounds of reasoning and experimentation) discovers distributed systems algorithms (e.g., GPU load balancing) that perform on par with those designed by world-renowned human experts in…
We built a Systems Researcher AI agent! Glia discovers novel distributed systems algorithms matching PhD-level experts in creativity & performance. We ran it on various networked systems problems and obtained publication-worthy results on each! Let me tell you how we did it 🧵
Bad news for the semiconductor industry. The proposed new USPTO rule would force patent holders to pay an annual fee ranging from 1% to 5% of a patent's assessed value, potentially in addition to current maintenance fees. - Highest volume: In 2023, for the second year in a row,…
High precision manipulation with tactile sensors. Figuring out the right representation for including tactile data is really important
Teaching robots sub-millimeter precision through spatially-grounded touch We introduce SaTA which anchors tactile to hand kinematics, enabling learning-based policies to match model-based precision. arxiv.org/abs/2510.14647
Cerebras beats Nvidia H100 but can it beat Blackwell? Blackwell inference endpoints are finally out and it’s fast. It runs GPT-OSS-120B at ~700 tokens/s, leapfrogging H100 and Groq. Cerebras clocked in at 3,000 TPS - still #1. Looking forward to Rubin!
had fun building a fake financial market with @extropic’s thermocomputing library thrml, using a modified Ising model (agents buy (+1) / sell (-1) via peer pressure + external hype) to simulate financial bubbles. prelim result: got fat tails and volatility clusters.
Lam Researchの先端パッケージングに関する動画ですが、101(ワンオーワン)の理由が知りたい Advanced Packaging Techniques (Semi 101) youtu.be/CAKWkOuoWrs?si…
youtube.com
YouTube
Advanced Packaging Techniques (Semi 101)
OpenArm: 7DOF bimanual humanoid arms for physical AI development in contact-rich tasks. Human-scale (63 cm reach), 6kg peak payload, 1kHz CAN-FD control, all HW/SW open-source. openarm.dev
Exciting big Kimi K2 Thinking release! More experts, fewer heads, and even more thinking!
🚀 Hello, Kimi K2 Thinking! The Open-Source Thinking Agent Model is here. 🔹 SOTA on HLE (44.9%) and BrowseComp (60.2%) 🔹 Executes up to 200 – 300 sequential tool calls without human interference 🔹 Excels in reasoning, agentic search, and coding 🔹 256K context window Built…
What if you could give an AI exactly where and what to draw—and have it nail both layout and style? Enter InstanceAssemble: a new Layout-to-Image method that uses instance-assembling attention to fuse bounding boxes, text, and even extra visual cues into pixel-perfect scenes.…
The paper shows how LLMs can judge product recommendations at scale without humans while staying reliable. It introduces ScalingEval, a multi agent setup that audits pairs, flags issues, and writes reports. Agents check complement patterns and tag problems like wrong size or…
United States Trends
- 1. #LingOrm1st_ImpactFANCON 723K posts
- 2. Talus Labs 25.3K posts
- 3. #BUNCHITA 1,427 posts
- 4. #KirbyAirRiders 1,983 posts
- 5. Frankenstein 84K posts
- 6. Giulia 15.9K posts
- 7. taylor york 9,224 posts
- 8. #SmackDown 49K posts
- 9. #River 4,980 posts
- 10. Ketanji Brown Jackson 4,919 posts
- 11. Tulane 4,512 posts
- 12. Justice Jackson 6,289 posts
- 13. Pluribus 31.7K posts
- 14. Aaron Gordon 5,444 posts
- 15. Russ 14.6K posts
- 16. Guillermo del Toro 26.4K posts
- 17. Tatis 2,287 posts
- 18. Connor Bedard 3,288 posts
- 19. Collar 17.2K posts
- 20. Supreme Court 183K posts
You might like
-
subho
@subhobrata1 -
エッジ・トランスレーション
@rose_miura -
wassname 🤔🧐
@wassname -
whjxnyzh
@whjxnyzhQAQ -
Anuj Dutt
@anujdutt92 -
Zach Bessinger
@ZachBessinger -
Mattia Verasani
@MatRazor -
Debi
@IanonsI -
Yi-Ting Shen
@DennisYTShen -
Maribel Monroy
@MaribelMonroy1 -
Harsh Kohli 🏴☠️
@HarshKohli09 -
pj_privatequity
@pj_privatequity -
Victor
@VCidFiguera -
Sagar
@sagarpath
Something went wrong.
Something went wrong.