wb_bitsbytes's profile picture. Charting new territories in digital realms!

Walker Bytes

@wb_bitsbytes

Charting new territories in digital realms!

Walker Bytes podał dalej

“AI is not distinct from humanity, but rather is a recent addition to a mutually interdependent superhuman entity we are all already part of.” It’s like humanity is building an external genome as a storehouse of common knowledge. nature.com/articles/d4158…


Walker Bytes podał dalej

Poured days into this poster, but it didn't get seen much, so I’m posting it here to give it a better home 🌟. #NeurIPS2025 If it’s up your alley, take a peek! 💛

breadli428's tweet image. Poured days into this poster, but it didn't get seen much, so I’m posting it here to give it a better home 🌟. #NeurIPS2025

If it’s up your alley, take a peek! 💛

🧠Model-Based RL shows promises but has seen limited success in real-world robotics. 🌎Introducing Robotic World Model, a black-box end-to-end neural dynamics model that bridges this gap, where policies are trained purely in imagination. @NeurIPSConf 🎯sites.google.com/view/roboticwo…



Walker Bytes podał dalej

Interested increased adaptivity and longer term memory in SSMs? Come by our poster #3412 today @ 4:30pm! "Bridging Expressivity and Scalability with Adaptive Unitary SSMs" Poster: neurips.cc/virtual/2025/l… Paper: arxiv.org/abs/2507.05238


Walker Bytes podał dalej

NEURIPS! I just arrived! Check out our group's papers and find me + my students to talk science! - CryoBoltz ❄️⚡️: Multiscale guidance of protein structure prediction with heterogeneous cryo-EM data @rishwanth_raghu arxiv.org/abs/2506.04490 - ChefNMR 🧑‍🍳⚛️: Atomic Diffusion…


Walker Bytes podał dalej

We are in the midst of a transformation phase We are knee-deep in a transformation, not with a bang, but with a quiet click and a subtle scroll. AI isn’t just reshaping the world, it’s reshaping us: how we think, what we value, even how we speak. And the strangest part? It’s…


Walker Bytes podał dalej

ComfyUI-Mesh2Motion Bring another popular open-source projects into ComfyUI, today it’s Mesh2Motion, from @scottpetrovic often called the open-source alternative to Mixamo. I spent some time integrating it into ComfyUI. It supports opening Mesh2Motion directly within ComfyUI,…


Walker Bytes podał dalej

Will computers ever match or surpass human-level intelligence — and, if so, how? go.nature.com/48kv5rM


Walker Bytes podał dalej

In our clinical trial, participants with geographic atrophy experienced something previously thought impossible: meaningful restoration of central vision. Learn more about our PRIMA retinal implant here: science.xyz/technologies/p…


Walker Bytes podał dalej

Congrats to the great #NeurIPS best paper (arxiv.org/pdf/2505.17638) on memorization in diffusion models. But note: producing non-memorized samples is not the same as generalization in the sense of learning structure or rules. Understanding generalization in DMs remains open.


Walker Bytes podał dalej

Robots need to be able to apply pressure and make contact with objects as needed in order to accomplish their tasks. From compliance to working safely around humans to whole-body manipulation of heavy objects, combining force and position control can dramatically expand the…


Walker Bytes podał dalej

🚨 Looking for a primer on the latest advances in joint neural-behavioral modeling? 🤩 Check out our new Nature Reviews Neuroscience piece! It was really an honor to write this 🫶 nature.com/articles/s4158…


Walker Bytes podał dalej

A statistical framework quantifies single-cell batch variation and recovers meaningful biological signals go.nature.com/4hZnzW5 rdcu.be/eRy6o


Walker Bytes podał dalej

Measuring Agents in Production @melissapan et al. investigate how AI agents are built and deployed across 26 industries, finding that 73% aim to boost productivity through automation while relying on simple, controllable methods over complex autonomy. 📝 arxiv.org/abs/2512.04123


Walker Bytes podał dalej

Congratulations @JesseJenkins , Camus, @Google @tylerhnorris on a fantastic paper! Great framing from @JigarShahDC

Ten tweet nie jest już dostępny.

Walker Bytes podał dalej

LLM token prices are collapsing fast, and the collapse is steepest at the top end. The least "intelligent" models get about 9× cheaper per year, mid-tier models drop about 40× per year, and the most capable models fall about 900× per year. Was same with "Moore’s Law, the best…

rohanpaul_ai's tweet image. LLM token prices are collapsing fast, and the collapse is steepest at the top end. 

The least "intelligent" models get about 9× cheaper per year, mid-tier models drop about 40× per year, and the most capable models fall about 900× per year.

Was same with "Moore’s Law, the best…

Walker Bytes podał dalej

Check out our #NeurIPS2025 paper Learning from Reward-Free Offline Data: A Case for Planning with Latent Dynamics Models Friday, afternoon poster session, Poster #212 Paper: latent-planning.github.io Joint work w/ @vlad_is_ai, Wancong Zhang, @kchonyc @randall_balestr @ylecun

What is the best way to train a general agent using offline state action trajectories without reward annotations? We test methods ranging from RL to optimal control, and find that planning with a latent dynamics model (PLDM) is a strong candidate. latent-planning.github.io



Walker Bytes podał dalej

A key dream of interpretability is building lie detectors. My team spent 6+ months on a project in this area, but unfortunately ran into a lot of issues... Our position piece describes why this was so hard, key problems future work must tackle, and advice for future researchers

NeelNanda5's tweet image. A key dream of interpretability is building lie detectors. My team spent 6+ months on a project in this area, but unfortunately ran into a lot of issues...

Our position piece describes why this was so hard, key problems future work must tackle, and advice for future researchers

can we detect when an AI system is being strategically deceptive without relying on behavioural evidence? we, on the GDM mechanistic interpretability team, spent many months working on this problem in our new position piece, we share what we learned:

bilalchughtai_'s tweet image. can we detect when an AI system is being strategically deceptive without relying on behavioural evidence?

we, on the GDM mechanistic interpretability team, spent many months working on this problem

in our new position piece, we share what we learned:


Walker Bytes podał dalej

MLX-LM has extreme alpha. It’s a unified inference and training framework that works across MLX and CUDA and constantly gets new features without losing modularity. It will only get better.

The latest mlx-lm is out and it has continuous batching with mlx_lm.server! Added by @angeloskath Check-out the video of 4 simultaneous requests running with Qwen3 30B on the same M2 Ultra:



Walker Bytes podał dalej

💡 With data center power becoming more constrained each year and the ever-growing importance of energy efficiency, we are committed to meeting this challenge. 🔋 With the new NVIDIA Power Profiles, conquer power constraints on your AI and HPC workloads and deliver up to 15%…

NVIDIADC's tweet image. 💡 With data center power becoming more constrained each year and the ever-growing importance of energy efficiency, we are committed to meeting this challenge.

🔋 With the new NVIDIA Power Profiles, conquer power constraints on your AI and HPC workloads and deliver up to 15%…

Walker Bytes podał dalej

Excited to share our #NeurIPS2025 work: NuCLR, a framework for learning neuron-level representations 🧠 These embeddings capture the biological identity of neurons and work out-of-the-box on new animals; no finetuning needed 💃 This offers some of the first evidence that…


To konto obecnie nikogo nie obserwuje

United States Trendy

Loading...

Something went wrong.


Something went wrong.