TT Nguyen @Machine1235

Author of GRPO-Zero library, PhD, AlphaZero in JAX, text-to-speech models, mechanistic interpretability, exploring neural nets' internal circuits... github.com/ntt123 Singapore Joined February 2012

Tweets

675
Followers

298
Following

714
Likes

3K

TT Nguyen @Machine1235

3 days ago

Anthropic should slow down and fix its services.

Anthony Morris ツ @amorriscode

3 days ago

working on a fix for issues with Opus 4.7 and Opus 4.8 status.claude.com

15 6 127 33K 16

0 0 0 28 0

View Details

ok i really like this model architecture, which somewhat proves my intuition that we don't really need a deep encoder for vision/audio modalities, the transformer backbone itself can process the image/audio tokens raw. The vision/audio encoder should be considered to play the same role as the "embedding" layer.

Google Gemma @googlegemma

5 days ago

Meet Gemma 4 12B! A unified, encoder-free multimodal model designed to bring high-performance intelligence directly to your laptop, and released under an Apache 2.0 license. Bridging the gap between edge efficiency and advanced reasoning. Here is what’s new with Gemma 4 12B: 👇

394 2K 12K 3.1M 5K

0 0 0 110 0

View Details

TT Nguyen @Machine1235

3 weeks ago

Another day of Google not quite reaching the AI capability frontier.

0 0 0 31 0

View Details

TT Nguyen @Machine1235

3 weeks ago

This isn't actually true, at least with the current version! (That said, Anthropic does offer a context editing feature in the API that lets you clear tool results and/or thinking blocks.) If you want to test it yourself, try this: 1 Ask Claude: "What is 123123123123 x 34534543564654645? No tools, just thinking." 2 Step away for 2 hours to make sure the KV cache gets flushed. 3 Come back and ask: "Give me a 10-word summary for every 200 words in your thinking block before the final answer."

0 0 0 231 0

View Details

TT Nguyen @Machine1235

3 weeks ago

They don't teach you this in school, but Transformers are RNNs. The KV cache is the hidden state.

François Fleuret @francoisfleuret

3 weeks ago

About "transformers are recurrent": teacher-forced discrete tokens cannot play a mind's melody.

10 1 63 15K 33

0 0 0 95 0

View Details

TT Nguyen @Machine1235

3 weeks ago

@Figure_robot This looks very impressive, but am I the only one seeing something off here? Do we really need a humanoid robot for this particular task? It feels like someone with a mechanical engineering degree could design a simple machine to do the same thing, right?

1 0 11 1K 0

View Details

TT Nguyen @Machine1235

3 weeks ago

In this clip, Dario was talking about *the latest model we release, Claude Opus 4.5* I've already seen a few cases where people post old clips of Dario for various reasons. I'm not commenting on whether Dario is purposely sending *fear-mongering* messages or if he honestly believes in this stuff. I just want to point out that some people are purposely posting old clips of Dario for engagement and/or other agendas.

3 0 7 2K 0

View Details

TT Nguyen @Machine1235

3 weeks ago

@tdietterich Someone needs to create a Claude skill to check for hallucinated references and other "meta comments from the LLM" 😅

0 0 0 330 0

View Details

TT Nguyen @Machine1235

4 weeks ago

Cancel your subscription and switch to Codex instead. You'll be doing others a favor too, freeing up compute for the people who actually need it.

Jun Song @jun_song

4 weeks ago

Best AI models now : Backend : GPT-5.5 Frontend : Kimi-K2.6 Agentic : MiMo-V2.5-Pro Best Price : Deepseek-V4-Flash Worst Price : Opus-4.7 Fast Research : Grok 4.3 Image : GPT-Image-2 Video : Seedance 2.0 (#1 for 3 months) Yes, Claude sucks.

171 246 4K 167K 2K

0 0 0 123 0

View Details

TT Nguyen @Machine1235

4 weeks ago

I am kind of tired of people's crazy takes on Anthropic. AI twitter is an upside-down world in which the honest people get ridiculed while the liars get celebrated.

0 0 0 48 0

View Details

TT Nguyen @Machine1235

4 weeks ago

Too much of the *hype* around Mythos is focused on cyber. But… Mythos is incredibly good at SWE as well. In fact, cyber is really just a side quest. I'd bet Mythos is great at other tasks too. And I don't think we're prepared for this.

0 0 0 74 0

View Details

TT Nguyen @Machine1235

4 weeks ago

If you want to listen to Claude’s Constitution in your favorite podcast app, I have created an RSS link for it at gist.githubusercontent.com/NTT123/173164a… Use your app’s Add podcast by URL / RSS URL option. On iOS Apple Podcasts: Library → Shows → Follow a Show by URL. On Android AntennaPod/Podcast Addict: add/import by RSS URL.

1 0 2 341 0

View Details

TT Nguyen @Machine1235

4 weeks ago

@yaroslavvb Much better developer experience (Python vs. C++). Much better compile time. Same performance as C++ CUDA.

0 0 0 433 0

View Details

TT Nguyen @Machine1235

4 weeks ago

github.com/NTT123/hill-cl… In Claude Code: /plugin marketplace add NTT123/hill-climbing-skills /plugin install hillclimb@hill-climbing-skills Then: "Run hillclimb on X. Maximize Y from baseline Z."

0 0 0 73 0

View Details

TT Nguyen @Machine1235

4 weeks ago

The dashboard is one self-contained HTML file. State + chart in one file you can email around. Here is a live example, mid-run on ResNet-18 / Flowers-102 fine-tune: ntt123.github.io/hill-climbing-…

1 0 0 199 0

View Details

TT Nguyen @Machine1235

4 weeks ago

I built a Claude Code plugin this weekend: turn any "chase a number" problem into a stateful hill-climbing loop. You give a metric and a verifier script. Claude does the iterating. State lives in a self-contained HTML dashboard. 🧵 below ↓

1 0 0 105 0

View Details

TT Nguyen @Machine1235

a month ago

I think looking at neural activity through a geometric lens is very promising. Here's another (simple) example from my earlier Logit Prisms work: when we look at digit embeddings inside an LLM, they form a nice heart-like <3 shape in 2D PCA.

Goodfire @GoodfireAI

a month ago

Neural networks might speak English, but they think in shapes. Understanding their rich *neural geometry* is key to understanding how they work – and to debugging and controlling them with precision. Starting today, we’re releasing a series of posts on this research agenda. 🧵

307 2K 11K 3.2M 9K

1 0 1 206 0

View Details

TT Nguyen @Machine1235

a month ago

One interesting thing I learned from @dwarkesh_sp and @reinerpope's recent lecture that few people mention is that the 1m and 1h caches are likely stored on Flash and HDD. That was a bit of a surprise to me. I thought the 1m cache was stored in HBM, which turns out to be wrong.