@ikyle.bsky.social
Reposted
Xogot Connect is live in the Godot Asset Library!
The easiest way to run and debug Godot games on iPhone and iPad:
• Deploy wirelessly over Wi-Fi
• Debug & Inspect nodes
• Live-test changes in realtime
• Built-in onscreen controls
Learn more:
blog.xogot.com/introducing-...
The easiest way to run and debug Godot games on iPhone and iPad:
• Deploy wirelessly over Wi-Fi
• Debug & Inspect nodes
• Live-test changes in realtime
• Built-in onscreen controls
Learn more:
blog.xogot.com/introducing-...
Introducing Xogot Connect — Remote Debugging for Godot on iPhone and iPad
Today we’re excited to announce the availability of Xogot Connect, a new Godot extension available in the Godot Asset Library. Xogot Connnect bridges your desktop Godot editor to Xogot on iOS enabling...
blog.xogot.com
November 12, 2025 at 3:34 PM
Xogot Connect is live in the Godot Asset Library!
The easiest way to run and debug Godot games on iPhone and iPad:
• Deploy wirelessly over Wi-Fi
• Debug & Inspect nodes
• Live-test changes in realtime
• Built-in onscreen controls
Learn more:
blog.xogot.com/introducing-...
The easiest way to run and debug Godot games on iPhone and iPad:
• Deploy wirelessly over Wi-Fi
• Debug & Inspect nodes
• Live-test changes in realtime
• Built-in onscreen controls
Learn more:
blog.xogot.com/introducing-...
Reposted
Someone at work just called the new Steam Machine a “GabeCube” and I can’t stop laughing.
steampowered.com/hardware
steampowered.com/hardware
Steam Hardware
The Steam Hardware family officially expands in early 2026.
steampowered.com
November 12, 2025 at 7:27 PM
Someone at work just called the new Steam Machine a “GabeCube” and I can’t stop laughing.
steampowered.com/hardware
steampowered.com/hardware
Reposted
If you’ve not seen them before, Jake’s interactive intro to CRDTs is absolutely fantastic.
jakelazaroff.com/words/an-int...
jakelazaroff.com/words/an-int...
June 14, 2025 at 8:59 PM
If you’ve not seen them before, Jake’s interactive intro to CRDTs is absolutely fantastic.
jakelazaroff.com/words/an-int...
jakelazaroff.com/words/an-int...
Reposted
Anthropic provides users with a best practices guide for Claude Code. Claude Code is a low-level tool without specific workflow restrictions, allowing users to directly access the Claude model.
www.anthropic.com/engineering/...
www.anthropic.com/engineering/...
Claude Code Best Practices
A blog post covering tips and tricks that have proven effective for using Claude Code across various codebases, languages, and environments.
www.anthropic.com
April 20, 2025 at 2:47 PM
Anthropic provides users with a best practices guide for Claude Code. Claude Code is a low-level tool without specific workflow restrictions, allowing users to directly access the Claude model.
www.anthropic.com/engineering/...
www.anthropic.com/engineering/...
Reposted
After 5 years in the making, I'm very happy to finally release Anime.js v4, my dream animation library:
🧩 New modular API
⚡️ Highly performant
🖱️ Scroll-linked animations
🫵 Draggables
↔️ Responsive
✨ Additive animations
🪽 WAAPI support
And so much more!
Check out the new website animejs.com!
🧩 New modular API
⚡️ Highly performant
🖱️ Scroll-linked animations
🫵 Draggables
↔️ Responsive
✨ Additive animations
🪽 WAAPI support
And so much more!
Check out the new website animejs.com!
April 3, 2025 at 3:32 PM
After 5 years in the making, I'm very happy to finally release Anime.js v4, my dream animation library:
🧩 New modular API
⚡️ Highly performant
🖱️ Scroll-linked animations
🫵 Draggables
↔️ Responsive
✨ Additive animations
🪽 WAAPI support
And so much more!
Check out the new website animejs.com!
🧩 New modular API
⚡️ Highly performant
🖱️ Scroll-linked animations
🫵 Draggables
↔️ Responsive
✨ Additive animations
🪽 WAAPI support
And so much more!
Check out the new website animejs.com!
Reposted
Wrote some notes on Aider's unconventional but kind of brilliant installation mechanism, which solves the "isolated Python virtual environment" problem by bundling uv as a dependency! simonwillison.net/2025/Mar/6/a...
![Provided you already have a Python install of version 3.8 or higher you can run this:
pip install aider-install && aider-install
The aider-install package itself depends on uv. When you run aider-install it executes the following Python code:
def install_aider():
try:
uv_bin = uv.find_uv_bin()
subprocess.check_call([
uv_bin, "tool", "install", "--force", "--python", "python3.12", "aider-chat@latest"
])
subprocess.check_call([uv_bin, "tool", "update-shell"])
except subprocess.CalledProcessError as e:
print(f"Failed to install aider: {e}")
sys.exit(1)](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:kft6lu4trxowqmter2b6vg6z/bafkreic44x3ul6bp2thefj2mykonzpoz7vypf2dg5thjlgvpexdur4tvqi@jpeg)
March 6, 2025 at 1:51 AM
Wrote some notes on Aider's unconventional but kind of brilliant installation mechanism, which solves the "isolated Python virtual environment" problem by bundling uv as a dependency! simonwillison.net/2025/Mar/6/a...
Reposted
A Waymo vehicle was driving in a 25mph zone in LA when an oncoming car swerved into its lane while speeding up to over 70mph… 3x the speed means 9x the destructive energy.
Waymo vehicle reacted safely.
Source: Dmitri Dolgov (Co-CEO at waymo)
Waymo vehicle reacted safely.
Source: Dmitri Dolgov (Co-CEO at waymo)
March 6, 2025 at 3:46 AM
A Waymo vehicle was driving in a 25mph zone in LA when an oncoming car swerved into its lane while speeding up to over 70mph… 3x the speed means 9x the destructive energy.
Waymo vehicle reacted safely.
Source: Dmitri Dolgov (Co-CEO at waymo)
Waymo vehicle reacted safely.
Source: Dmitri Dolgov (Co-CEO at waymo)
Reposted
Alibaba has open-sourced QwQ-32B, a cutting-edge large language model with 32 billion parameters, designed to enhance reasoning capabilities through the power of Reinforcement Learning (RL).
Try it here: chat.qwen.ai
Try it here: chat.qwen.ai
March 6, 2025 at 6:01 AM
Alibaba has open-sourced QwQ-32B, a cutting-edge large language model with 32 billion parameters, designed to enhance reasoning capabilities through the power of Reinforcement Learning (RL).
Try it here: chat.qwen.ai
Try it here: chat.qwen.ai
Reposted
Alibaba's Babel - Open Multilingual Large Language Models Serving Over 90% of Global Speakers
Project: babel-llm.github.io/babel-llm/ with the links to
- Paper
- GitHub
- Models
Project: babel-llm.github.io/babel-llm/ with the links to
- Paper
- GitHub
- Models
March 6, 2025 at 6:37 AM
Alibaba's Babel - Open Multilingual Large Language Models Serving Over 90% of Global Speakers
Project: babel-llm.github.io/babel-llm/ with the links to
- Paper
- GitHub
- Models
Project: babel-llm.github.io/babel-llm/ with the links to
- Paper
- GitHub
- Models
Reposted
New blog post.
bartoszmilewski.com/2025/03/06/u...
bartoszmilewski.com/2025/03/06/u...
Understanding Attention in LLMs
There are many excellent AI papers and tutorials that explain the attention pattern in Large Language Models. But this essentially simple pattern is often obscured by implementation details and opt…
bartoszmilewski.com
March 6, 2025 at 7:03 PM
New blog post.
bartoszmilewski.com/2025/03/06/u...
bartoszmilewski.com/2025/03/06/u...
Reposted
Published some notes on S1, a new reasoning model that was fine-tuned from Qwen 2.5 32B on just 1,000 examples and $6 of compute
Includes notes on running it with Ollama and exploring those 1,000 examples with Datasette Lite simonwillison.net/2025/Feb/5/s...
Includes notes on running it with Ollama and exploring those 1,000 examples with Datasette Lite simonwillison.net/2025/Feb/5/s...
S1: The $6 R1 Competitor?
Tim Kellogg shares his notes on a new paper, [s1: Simple test-time scaling](https://arxiv.org/abs/2501.19393), which describes an inference-scaling model fine-tuned on top of Qwen2.5-32B-Instruct for ...
simonwillison.net
February 5, 2025 at 8:03 PM
Published some notes on S1, a new reasoning model that was fine-tuned from Qwen 2.5 32B on just 1,000 examples and $6 of compute
Includes notes on running it with Ollama and exploring those 1,000 examples with Datasette Lite simonwillison.net/2025/Feb/5/s...
Includes notes on running it with Ollama and exploring those 1,000 examples with Datasette Lite simonwillison.net/2025/Feb/5/s...
Reposted
Deep Dive into LLMs like ChatGPT by Andrej Karpathy
(3 hours 31 minutes)
Audience: General audience deep dive into LLM AI
Covers:
1. pretraining
2. supervised finetuning
3. reinforcement learning
Link: www.youtube.com/watch?v=7xTG...
(3 hours 31 minutes)
Audience: General audience deep dive into LLM AI
Covers:
1. pretraining
2. supervised finetuning
3. reinforcement learning
Link: www.youtube.com/watch?v=7xTG...
Deep Dive into LLMs like ChatGPT
YouTube video by Andrej Karpathy
www.youtube.com
February 5, 2025 at 8:05 PM
Deep Dive into LLMs like ChatGPT by Andrej Karpathy
(3 hours 31 minutes)
Audience: General audience deep dive into LLM AI
Covers:
1. pretraining
2. supervised finetuning
3. reinforcement learning
Link: www.youtube.com/watch?v=7xTG...
(3 hours 31 minutes)
Audience: General audience deep dive into LLM AI
Covers:
1. pretraining
2. supervised finetuning
3. reinforcement learning
Link: www.youtube.com/watch?v=7xTG...
Reposted
<table> highlighting with CSS :has() 💡
td:has(~ td:hover), /* previous sibling cells */
table:has(td:nth-of-type(3):hover) /* column cells */
tr:not(:first-of-type):has(~ tr:hover)
td:nth-of-type(3) {
background: var(--highlighted);
}
td:has(~ td:hover), /* previous sibling cells */
table:has(td:nth-of-type(3):hover) /* column cells */
tr:not(:first-of-type):has(~ tr:hover)
td:nth-of-type(3) {
background: var(--highlighted);
}
February 5, 2025 at 9:07 PM
<table> highlighting with CSS :has() 💡
td:has(~ td:hover), /* previous sibling cells */
table:has(td:nth-of-type(3):hover) /* column cells */
tr:not(:first-of-type):has(~ tr:hover)
td:nth-of-type(3) {
background: var(--highlighted);
}
td:has(~ td:hover), /* previous sibling cells */
table:has(td:nth-of-type(3):hover) /* column cells */
tr:not(:first-of-type):has(~ tr:hover)
td:nth-of-type(3) {
background: var(--highlighted);
}
Reposted
100,277 tokens – The vocabulary size of GPT-4’s tokenizer, showcasing the complexity of tokenization.
$40,000 (2019) → $100 (2024) – The reduction in cost to train a GPT-2 equivalent model, illustrating improvements in efficiency and hardware advancements.
$40,000 (2019) → $100 (2024) – The reduction in cost to train a GPT-2 equivalent model, illustrating improvements in efficiency and hardware advancements.
February 6, 2025 at 2:28 AM
100,277 tokens – The vocabulary size of GPT-4’s tokenizer, showcasing the complexity of tokenization.
$40,000 (2019) → $100 (2024) – The reduction in cost to train a GPT-2 equivalent model, illustrating improvements in efficiency and hardware advancements.
$40,000 (2019) → $100 (2024) – The reduction in cost to train a GPT-2 equivalent model, illustrating improvements in efficiency and hardware advancements.
Reposted
Paper: Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languages by one of the author - @jannikbrinkmann.bsky.social ( arxiv.org/abs/2501.06346 )
Repo: github.com/jannik-brink...
Post on X: x.com/jannikbrinkm...
Repo: github.com/jannik-brink...
Post on X: x.com/jannikbrinkm...
Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languages
Human bilinguals often use similar brain regions to process multiple languages, depending on when they learned their second language and their proficiency. In large language models (LLMs), how are mul...
arxiv.org
February 6, 2025 at 7:19 AM
Paper: Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languages by one of the author - @jannikbrinkmann.bsky.social ( arxiv.org/abs/2501.06346 )
Repo: github.com/jannik-brink...
Post on X: x.com/jannikbrinkm...
Repo: github.com/jannik-brink...
Post on X: x.com/jannikbrinkm...
Reposted
Why do LLMs trained on over 90% English text perform so well in non-English languages?
They find that they learn to share highly abstract grammatical concept representations, even across unrelated languages!
They find that they learn to share highly abstract grammatical concept representations, even across unrelated languages!
February 6, 2025 at 7:19 AM
Why do LLMs trained on over 90% English text perform so well in non-English languages?
They find that they learn to share highly abstract grammatical concept representations, even across unrelated languages!
They find that they learn to share highly abstract grammatical concept representations, even across unrelated languages!
Reposted
Llama2’s internal “lingua franca” is not English, but concepts - and, crucially, concepts that are biased toward English. Hence, English could still be seen as an “internal language”, but in a semantic, rather than a purely lexical.
Paper: arxiv.org/abs/2402.10588
Post on X: x.com/cervisiarius...
Paper: arxiv.org/abs/2402.10588
Post on X: x.com/cervisiarius...
February 6, 2025 at 7:26 AM
Llama2’s internal “lingua franca” is not English, but concepts - and, crucially, concepts that are biased toward English. Hence, English could still be seen as an “internal language”, but in a semantic, rather than a purely lexical.
Paper: arxiv.org/abs/2402.10588
Post on X: x.com/cervisiarius...
Paper: arxiv.org/abs/2402.10588
Post on X: x.com/cervisiarius...
Reposted
Andrej's re-recorded video "Intro to LLMs" is worth watching. The newly made 3-hour version is aimed at a general audience, covering the complete training process of LLM model development, the mental models for understanding their "psychology," and how to maximize their use in practical applications
Deep Dive into LLMs like ChatGPT
YouTube video by Andrej Karpathy
www.youtube.com
February 6, 2025 at 4:10 PM
Andrej's re-recorded video "Intro to LLMs" is worth watching. The newly made 3-hour version is aimed at a general audience, covering the complete training process of LLM model development, the mental models for understanding their "psychology," and how to maximize their use in practical applications
Reposted
Paper: arxiv.org/abs/2502.03382
Samples: hf.co/spaces/kyuta...
Inference code: github.com/kyutai-labs/...
Models: huggingface.co/kyutai
Post on X: x.com/kyutai_labs/...
Samples: hf.co/spaces/kyuta...
Inference code: github.com/kyutai-labs/...
Models: huggingface.co/kyutai
Post on X: x.com/kyutai_labs/...
High-Fidelity Simultaneous Speech-To-Speech Translation
We introduce Hibiki, a decoder-only model for simultaneous speech translation. Hibiki leverages a multistream language model to synchronously process source and target speech, and jointly produces tex...
arxiv.org
February 7, 2025 at 7:41 AM
Paper: arxiv.org/abs/2502.03382
Samples: hf.co/spaces/kyuta...
Inference code: github.com/kyutai-labs/...
Models: huggingface.co/kyutai
Post on X: x.com/kyutai_labs/...
Samples: hf.co/spaces/kyuta...
Inference code: github.com/kyutai-labs/...
Models: huggingface.co/kyutai
Post on X: x.com/kyutai_labs/...
Reposted
@kyutai-labs.bsky.social releases Hibiki, a simultaneous speech-to-speech translation model, currently supporting 🇫🇷➡️🇬🇧.
Hibiki produces spoken and text translations of the input speech in real-time, while preserving the speaker’s voice and optimally adapting its pace based on the semantic content.
Hibiki produces spoken and text translations of the input speech in real-time, while preserving the speaker’s voice and optimally adapting its pace based on the semantic content.
February 7, 2025 at 7:41 AM
@kyutai-labs.bsky.social releases Hibiki, a simultaneous speech-to-speech translation model, currently supporting 🇫🇷➡️🇬🇧.
Hibiki produces spoken and text translations of the input speech in real-time, while preserving the speaker’s voice and optimally adapting its pace based on the semantic content.
Hibiki produces spoken and text translations of the input speech in real-time, while preserving the speaker’s voice and optimally adapting its pace based on the semantic content.
Reposted
We did it! Kokoro TTS (v1.0) can now run 100% locally in your browser w/ WebGPU acceleration. Real-time text-to-speech without a server. ⚡️
Generate 10 seconds of speech in ~1 second for $0.
What will you build? 🔥
Generate 10 seconds of speech in ~1 second for $0.
What will you build? 🔥
February 7, 2025 at 5:03 PM
We did it! Kokoro TTS (v1.0) can now run 100% locally in your browser w/ WebGPU acceleration. Real-time text-to-speech without a server. ⚡️
Generate 10 seconds of speech in ~1 second for $0.
What will you build? 🔥
Generate 10 seconds of speech in ~1 second for $0.
What will you build? 🔥
Reposted
To kick off using Bluesky: our new dataset called Oxford Spires. Synchronised, multi-color cameras and lidar in multiple Oxford colleges. Ground-Truth highly accurate 3D maps from tripod scanners. The ideal basis for NeRF/3DGS SLAM research.
dynamic.robots.ox.ac.uk/datasets/oxf...
dynamic.robots.ox.ac.uk/datasets/oxf...
December 9, 2024 at 10:29 AM
To kick off using Bluesky: our new dataset called Oxford Spires. Synchronised, multi-color cameras and lidar in multiple Oxford colleges. Ground-Truth highly accurate 3D maps from tripod scanners. The ideal basis for NeRF/3DGS SLAM research.
dynamic.robots.ox.ac.uk/datasets/oxf...
dynamic.robots.ox.ac.uk/datasets/oxf...
Reposted
Demo: playground.zyphra.com/audio
Blog: zyphra.com/post/beta-re...
Models:
huggingface.co/Zyphra/Zonos... and huggingface.co/Zyphra/Zonos...
Repo: github.com/Zyphra/Zonos
Blog: zyphra.com/post/beta-re...
Models:
huggingface.co/Zyphra/Zonos... and huggingface.co/Zyphra/Zonos...
Repo: github.com/Zyphra/Zonos
Zyphra Maia
playground.zyphra.com
February 10, 2025 at 6:44 PM
Reposted
Zyphra beta releases Zonos, a highly expressive TTS model with high fidelity voice cloning.
They release both transformer and SSM-hybrid models under an Apache 2.0 license.
They release both transformer and SSM-hybrid models under an Apache 2.0 license.
February 10, 2025 at 6:44 PM
Zyphra beta releases Zonos, a highly expressive TTS model with high fidelity voice cloning.
They release both transformer and SSM-hybrid models under an Apache 2.0 license.
They release both transformer and SSM-hybrid models under an Apache 2.0 license.