Delphine @delphinel

Paris, France Joined December 2008

Tweets

2K
Followers

896
Following

1K
Likes

322

François Chollet @fchollet

2 months ago

We were able to reproduce the strong findings of the HRM paper on ARC-AGI-1. Further, we ran a series of ablation experiments to get to the bottom of what's behind it. Key findings: 1. The HRM model architecture itself (the centerpiece of the paper) is not an important factor.…

46 297 3K 350K 2K

eigenron @eigenron

2 months ago

i remember having a 1 hour discussion post class time with my stats professor about why we divide by n-1 instead of n for the sample variance, after i first learnt about this in my stat analysis class. the reason is shockingly elegant but still a little abstract unless you've…

36 154 2K 157K 2K

Download Image

François Chollet @fchollet

2 months ago

GenAI isn't just a technology; it's an informational pollutant—a pervasive cognitive smog that touches and corrupts every aspect of the Internet. It's not just a productivity tool; it's a kind of digital acid rain, silently eroding the value of all information. Every image is no…

478 1K 7K 671K 2K

Leonie @helloiamleonie

2 months ago

Apple just released Embedding Atlas: An open-source visualization tool for your embeddings. I just gave it a quick spin with some data stored in my vector database. These are my first impressions: - Nice exploration UX with hover and tool tip for single data points - Shows you…

8 92 604 48K 656

Download Image

Lili @lchen915

2 months ago

Self-Questioning Language Models: LLMs that learn to generate their own questions and answers via asymmetric self-play RL. There is no external training data – the only input is a single prompt specifying the topic.

25 185 1K 142K 1K

Download Image

❄️Andrew Zhao❄️ @_AndrewZhao

5 months ago

❄️Introducing Absolute Zero Reasoner: Our reasoner learns to both propose tasks that maximize learnability and improve reasoning by solving them, entirely through self-play—with no external data! It overall outperforms other "zero" models in math & coding domains. 🧵 1/

60 341 2K 515K 2K

Download Image

Shruti @heyshrutimishra

2 months ago

This paper didn’t go viral but it should have. A tiny AI model called HRM just beat Claude 3.5 and Gemini. It doesn’t even use tokens. They said it was just a research preview. But it might be the first real shot at AGI. Here’s what really happened and why OpenAI should be…

350 1K 10K 1.3M 12K

Download Image

Alex Prompter @alex_prompter

2 months ago

what are large language models actually doing? i read the 2025 textbook "Foundations of Large Language Models" by tong xiao and jingbo zhu and for the first time, i truly understood how they work. here’s everything you need to know about llms in 3 minutes↓

77 940 7K 1.1M 18K

Download Image

Santiago @svpino

4 months ago

Everything MCP offers in a nutshell: Let's start with an MCP Server. An MCP Server can expose Tools, Resources, and Prompts: 1. Tools: These allow your application or agent to connect to external services (for example, to connect to an API). The model decides which tools to…

20 102 730 74K 1K

Download Image

Loic @LoicReco

3 months ago

MCP is redefining what AI can do, and 95% of builders haven't caught on. I've built agents that run my calendar, organize files, and manage tasks. All autonomous. The results changed how I think about automation. I'm breaking down exactly how MCP works (with code + examples):

62 367 3K 355K 5K

Download Image

Tulsi Soni @shedntcare_

4 months ago

Anthropic just opened the vault The best free masterclass on prompt engineering is here — and it’s a game-changer

90 515 4K 447K 8K

Download Image

cat @_catwu

4 months ago

Since we originally built Claude Code as an internal tool, we've heard a ton of questions about how our teams use it at Anthropic. Here’s an inside look on how our teams—from product engineering, to growth marketing, to legal—use Claude Code:

63 437 4K 833K 7K

Download Image

Aadit Sheth @aaditsh

4 months ago

This is the ultimate guide to Fine-Tuning LLMs

23 252 2K 192K 3K

Download Image

Leonie @helloiamleonie

4 months ago

Level 1: Keyword-based retrieval Level 2: Semantic-based retrieval Level 3: Reasoning-based retrieval 𝗪𝗵𝗮𝘁 𝗶𝘀 𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹?I It is the ability to find relevant information that requires logical thinking to connect your query…

21 296 2K 213K 3K

Download Image

Chris Hayes @chrislhayes

5 months ago

For over 30 years, the NIH has been running one of the biggest studies in the world on women’s postmenopausal health. It’s produced enormously significant findings. @POTUS just ended it. Would be curious to hear the rational from @elonmusk and @JDVance youtu.be/z7Hy_NSZgWo?si…

949 4K 9K 830K 655

Amy Diehl, Ph.D. @amydiehl

5 months ago

The NIH Women’s Health Initiative enrolled 10,000's of participants in trials of medications & tracked health of thousands for 3+ decades. “This trial taught us an immense amount about prevention of disease in women." The Trump admin is cancelling it.science.org/content/articl…

11 197 336 28K 27

Women in Research IE @WomenInResIE

5 months ago

“It’s a sad day indeed for women’s health research” NIH cancels its first and largest study centered on women | Science | AAAS science.org/content/articl…

0 9 11 1K 0

Delphine @delphinel

7 months ago

kpbs.org/news/economy/2… Researchers say their work is now at risk if it contains language deemed problematic by the White House, including the word “women.“

0 0 0 31 0

Andrej Karpathy @karpathy

9 months ago

DeepSeek (Chinese AI co) making it look easy today with an open weights release of a frontier-grade LLM trained on a joke of a budget (2048 GPUs for 2 months, $6M). For reference, this level of capability is supposed to require clusters of closer to 16K GPUs, the ones being…