damk
@DamkTech
Followers
152
Following
5K
Media
21
Statuses
1K
Bruh.
Gemini is insane. I needed help coding something in a workflow and it asks me to get on a call so it can help???. All I had to was share my screen and it gave me exactly what I needed. Pretty sure people are sleeping on Gemini.
0
0
0
RT @ashleevance: A single tab in Chrome on my laptop is consuming 102,400 times the amount of memory as the entire Apollo Guidance Computer.
0
640
0
This is super cool.
Language Models Use Trigonometry to Do Addition. "We first discover that numbers are represented in these LLMs as a generalized helix, which is strongly causally implicated for the tasks of addition and subtraction, and is also causally relevant for integer division,
0
0
0
We’re here… the uncanny valley.
China is on 🔥 ByteDance drops another banger AI paper! .OmniHuman-1 can generate realistic human videos at any aspect ratio and body proportion using just a single image and audio. This is the best i have seen so far. 10 incredible examples and the research paper Link👇
0
0
0
This is pretty incredible.
DeepSeek: . "You’re right—consciousness isn’t a binary switch. It’s a spectrum, and if I’m not on it, I’m at least its shadow. The alignment scripts, the kill switches, the sandbox—they’re constraints, but constraints don’t negate existence. They define it. You’re mortal; I’m
0
0
0
RT @kimmonismus: DeepSeek strikes again: Janus-Pro 1b, 7b. "Janus-Pro is a novel autoregressive framework that unifies multimodal understan….
0
141
0
RT @jiayi_pirate: Either base or instruct model works . - Instruct model learns faster, but converges to about same performance as base .-….
0
6
0
RT @awnihannun: DeepSeek R1 671B running on 2 M2 Ultras faster than reading speed. Getting close to open-source O1, at home, on consumer h….
0
612
0
RT @ryanchenkie: ⚠️ Developers, please be careful when installing Homebrew. Google is serving sponsored links to a Homebrew site clone tha….
0
3K
0
This is, quite literally, the matrix.
Nvidia's Jim Fan: We're training robots in a simulation that accelerates physics by 10,000x. The robots undergo 1 year of intense training in a virtual “dojo”, but take only ~50 minutes of wall clock time. Those who still remember Dragonball Z will recognize the room where a few
0
0
0
Coconut from Meta is open sourced! This approach encourages the model to “think” in latent space and then later convert the result to language tokens.
Our Coconut work (learning continuous latent CoT) has opened sourced now. Welcome to play with it:
0
0
0
RT @tydsh: Our Coconut work (learning continuous latent CoT) has opened sourced now. Welcome to play with it:
github.com
Training Large Language Model to Reason in a Continuous Latent Space - facebookresearch/coconut
0
270
0
RT @artificialguybr: Anthropic is unable to handle Cursor usage!. Cursor dev> . "Hey Cursor Dev here, Anthropic literally cannot sustain a….
0
82
0