DeepLearn007's profile picture. Imtiaz Adam CS #AI Postgrad |#Tech #Strategy #MachineLearning #DeepLearning | #RL #Agentic | #LLM Liberal | #GenAI| MBA alum @morganstanley @LBS @Columbia_Biz

AI

@DeepLearn007

Imtiaz Adam CS #AI Postgrad |#Tech #Strategy #MachineLearning #DeepLearning | #RL #Agentic | #LLM Liberal | #GenAI| MBA alum @morganstanley @LBS @Columbia_Biz

AI reposted

So DeepSeek-Math-V2. It could be subtitled: "how to train better verifiers?" and the bulk of it is simply… better data work and synth pipelines (even if all models are trained with RL). DeepSeek further distances itself from the initial promises of spontaneous…

Ok. This calls for a one hour close reading with a cup of hot chocolate.

Dorialexander's tweet image. Ok. This calls for a one hour close reading with a cup of hot chocolate.


AI reposted

AgiBOT’s humanoid robot A2 just achieved a Guinness World Record. It walked 106KM from Suzhou’s Jinji Lake all the way to the Shanghai Bund, earning the title of “Longest Distance Walked by a Humanoid Robot.” A2 ran non-stop for 3 days and 3 nights, powered by hot-swappable…

Between factories and homes, Shanghai-based AGIBOT Robotics is deploying the AGIBOT A2 in large quantities in commercial service scenarios, such as shopping malls, showrooms, and auto dealerships. They mass-produced 1,000 units earlier this year and plan to produce 5,000 this…



AI reposted

Researchers at the Hong Kong University of Science and Technology, in collaboration with Unitree Robotics, have unveiled the world’s first full-size humanoid robot that can autonomously play interactive basketball with a human.


AI reposted

New milestone unlocked! The hardest working robot in the industry has now moved more than 100,000 totes for @GXOLogistics. Read more about this historic marker. ⬇️ agilityrobotics.com/content/digit-…


AI reposted

A Chinese research team has developed a modular, flexible lower-limb exoskeleton robot with broad potential applications, including mountaineering, smart elderly care, medical rehabilitation and outdoor sports. #GLOBALink


AI reposted

Treventus scan robot processes up to 2500 pages per hour.


AI reposted

XSTO Biped Robot – Bionic Leg Technology for All-Terrain Mobility XSTO Mobility specializes in the development of advanced mobility solutions, including power chairs, stair-climbing wheelchairs, and bipedal robotic systems. The company’s bionic leg technology is designed to…


AI reposted

#A neuromorphic chip now interprets brain-like neural connectivity in real time, enabling up to 20,000 times faster processing for advanced #Brain-computerInterface and AI applications. doi.org/hbb5f3 techxplore.com/news/2025-11-b…


AI reposted

Neuromorphic vision with optical synapses Most machine-vision systems today still follow a von Neumann pipeline: light is captured by a sensor, digitized, shipped to memory, and then processed by a compute unit. The problem is not the algorithms—it’s the movement of data.…

bravo_abad's tweet image. Neuromorphic vision with optical synapses

Most machine-vision systems today still follow a von Neumann pipeline: light is captured by a sensor, digitized, shipped to memory, and then processed by a compute unit. The problem is not the algorithms—it’s the movement of data.…

AI reposted

Neuromorphic chips could solve AI’s energy crisis by fusing memory and computation, creating ultra-efficient, brain-like chips. policycircle.org/opinion/neurom…


AI reposted

An amazing blog post dropped on @huggingface explaining how today's LLM inference engines like @vllm_project work! The concept of "continuous batching" is explained, along with KV-caching, attention masking, chunked prefill, and decoding. Continuous batching is the idea of…

NielsRogge's tweet image. An amazing blog post dropped on @huggingface explaining how today's LLM inference engines like @vllm_project work!

The concept of "continuous batching" is explained, along with KV-caching, attention masking, chunked prefill, and decoding.

Continuous batching is the idea of…

AI reposted

LLM Cheat Sheet Source : @goyalshaliniuk

PythonPr's tweet image. LLM Cheat Sheet
Source : @goyalshaliniuk

AI reposted

Nano Banana Pro can do essentially perfect text, which means it can do slides - paired with Kling transitions, it's an insanely cool new format for how to do presentations here's a tutorial on how to do it all with an agent:


AI reposted

"Sir, Deepseek just dropped DeepSeek-Math-v2 and it beats Gemini DeepThink on IMO ProofBench and CNML"

askOkara's tweet image. "Sir, Deepseek just dropped DeepSeek-Math-v2 and it beats Gemini DeepThink on IMO ProofBench and CNML"

AI reposted

Opus 4.5 can beat Gemini 3 in frontend Run these 2 commands in Claude Code: 1. /plugin marketplace add anthropics/claude-code 2. /plugin install frontend-design@claude-code-plugins try it

donvito's tweet image. Opus 4.5 can beat Gemini 3 in frontend      

Run these 2 commands in Claude Code:     

1. /plugin marketplace add anthropics/claude-code    
2. /plugin install frontend-design@claude-code-plugins  

try it

AI reposted

All done with my new course on Deep Representation Learning this semester. All lecture slides and video recordings are now available at the book website: ma-lab-berkeley.github.io/deep-represent… We believe, with the book, they help students clarify basic concepts and principles of Intelligence.

YiMaTweets's tweet image. All done with my new course on Deep Representation Learning this semester. All lecture slides and video recordings are now available at the book website: ma-lab-berkeley.github.io/deep-represent… We believe, with the book, they help students clarify basic concepts and principles of Intelligence.

AI reposted

Something about honeybees saving the girls is so poetic

Australian scientists found that honeybee venom can destroy aggressive breast cancer cells in lab tests, a discovery that could inspire future treatments

sciencegirl's tweet image. Australian scientists found that honeybee venom can destroy aggressive breast cancer cells in lab tests, a discovery that could inspire future treatments


AI reposted

“LLMS all the way down” with a MoE they were able to achieve above human baseline on Arc AGI 2 by building on top of GPT 5.1 and Gemini 3 Pro.

chatgpt21's tweet image. “LLMS all the way down” with a MoE they were able to achieve above human baseline on Arc AGI 2 by building on top of GPT 5.1 and Gemini 3 Pro.
chatgpt21's tweet image. “LLMS all the way down” with a MoE they were able to achieve above human baseline on Arc AGI 2 by building on top of GPT 5.1 and Gemini 3 Pro.

Jesus, superhuman performance on ARCv2 now at around 50 dollars per task? 3 significant advances in the last week alone.

MLStreetTalk's tweet image. Jesus, superhuman performance on ARCv2 now at around 50 dollars per task? 3 significant advances in the last week alone.


AI reposted

We’re working with the Poetiq team to verify this score Once verified, we’ll publish results along with datapoints

Jesus, superhuman performance on ARCv2 now at around 50 dollars per task? 3 significant advances in the last week alone.

MLStreetTalk's tweet image. Jesus, superhuman performance on ARCv2 now at around 50 dollars per task? 3 significant advances in the last week alone.


AI reposted

Is more intelligence always more expensive? Not necessarily. Introducing Poetiq. We’ve established a new SOTA and Pareto frontier on @arcprize using Gemini 3 and GPT-5.1.

poetiq_ai's tweet image. Is more intelligence always more expensive? Not necessarily.

Introducing Poetiq. We’ve established a new SOTA and Pareto frontier on @arcprize using Gemini 3 and GPT-5.1.

Loading...

Something went wrong.


Something went wrong.