Salmone Fall

@LayeSeckSeckFCB J'aimerais bien connaitre le pourcentage de guinéens

‼️Officiel | Selon l’ANSD (Agence Nationale de la Statistique et de la Démographie). 🔹Dans l’Annuaire sur la Population du Sénégal 2025, la population estimée est exactement de 19 075 959 habitants. 🔸Répartition par sexe : 🙋‍♂️ Hommes : 9 670 361 (environ 50,7 %) 👩‍🦰 Femmes : 9 405 598 (environ 49,3 %) 📌Les hommes sont donc légèrement plus nombreux que les femmes au niveau national (environ 103 hommes pour 100 femmes). 🇸🇳❤️🏛️

@Cisse_Maps Business bu haram ni rek la meuna démé du soxor deh depigmentation bakhul

Nouvelle affaire qui secoue le monde des affaires au Sénégal. La société Diodio Glow Skin, spécialisée dans la vente et la transformation de produits cosmétiques, serait confrontée à un important dossier de fraude interne. Selon des informations relayées par Senego, un montant de plus de 200 millions FCFA aurait été soustrait par une partie de ses employés. Une affaire qui suscite déjà de nombreuses réactions. Le dossier transmis à la justice Toujours selon les mêmes informations, le dossier aurait déjà été transféré à la justice afin que toute la lumière soit faite sur cette affaire. FD média

@AyaSansFiltre Malheureusement on est pas la cible des femmes qui préfèrent les "bad boy" après ce sont les mêmes qui vont venir se plaindre dur ce réseau dingue 😂

Y a des garçons… ils respirent la bonté et la bienveillance, Ça se voit, ils feront de bons maris et de bons pères.

@BsmDiarra Perso j'ai vu beaucoup de divorces dans mon entourage qui ne m'ont pas surpris. Les gens ne se marient plus pour les bonnes raisons

Le nombre de divorces que je vois ces temps-ci me fait très peur!💔 té diomboul ken en plus

@LatyrDiouff La triste vérité que les gens ne veulent pas entendre c'est que les personnes ne se marient pas souvent pour les bonnes raisons entre xémém, argent, changement de statut social ce qui est très fragile.

Divorcer nak méttina deh… gnima guiss niou traverser, souffrir nagne lou metti… surtout djiguéne gni🥲 Gorgorlou lene wayy mou wagnékou

@linaa_ai Claude

I’m convinced: Claude is the most powerful AI tool for making money right now. If you use it to create digital assets today, you could make an extra $10,000/month. I compiled the exact prompts I use into a 53-page PDF. Usually, I'd charge $199 for this, but today I'm giving it away 100% FREE Like + comment 'Claude' & I'll DM it to you Must follow me to get DM. ⏳ Taking this down in 24 hours.

PhD students doing a literature review ask: have I read enough? That is the wrong question. The right question is: have I searched in the right places?

This paper from Harvard and MIT quietly answers the most important AI question nobody benchmarks properly: Can LLMs actually discover science, or are they just good at talking about it? The paper is called “Evaluating Large Language Models in Scientific Discovery”, and instead of asking models trivia questions, it tests something much harder: Can models form hypotheses, design experiments, interpret results, and update beliefs like real scientists? Here’s what the authors did differently 👇 • They evaluate LLMs across the full discovery loop hypothesis → experiment → observation → revision • Tasks span biology, chemistry, and physics, not toy puzzles • Models must work with incomplete data, noisy results, and false leads • Success is measured by scientific progress, not fluency or confidence What they found is sobering. LLMs are decent at suggesting hypotheses, but brittle at everything that follows. ✓ They overfit to surface patterns ✓ They struggle to abandon bad hypotheses even when evidence contradicts them ✓ They confuse correlation for causation ✓ They hallucinate explanations when experiments fail ✓ They optimize for plausibility, not truth Most striking result: `High benchmark scores do not correlate with scientific discovery ability.` Some top models that dominate standard reasoning tests completely fail when forced to run iterative experiments and update theories. Why this matters: Real science is not one-shot reasoning. It’s feedback, failure, revision, and restraint. LLMs today: • Talk like scientists • Write like scientists • But don’t think like scientists yet The paper’s core takeaway: Scientific intelligence is not language intelligence. It requires memory, hypothesis tracking, causal reasoning, and the ability to say “I was wrong.” Until models can reliably do that, claims about “AI scientists” are mostly premature. This paper doesn’t hype AI. It defines the gap we still need to close. And that’s exactly why it’s important.

I learned how transformers actually work in 22 minutes this evening. Not from a video or a paper. From talking out loud. I've tried to understand attention maybe six times. Andrej's video. The Illustrated Transformer post. Three different chatbots. Each time I'd get to paragraph four of someone's explanation and quietly start scrolling instead of reading. So at @emergentlabs vibecon this weekend, we built the thing I actually needed. Slate - a voice-first AI tutor that teaches you anything and everything you want, in under 30 minutes, while drawing it on a slate. You tell it a topic. You talk to it. It talks back warm, patient, two sentences at a time. And while it's explaining, it's drawing. Mermaid diagrams. Code with syntax highlighting. KaTeX equations. Chalk-on-blackboard sketches. It takes notes on the side that you can download as markdown. What it replaces, for me: • 3 YouTube tabs → one 20-minute conversation • A 40-page PDF → a diagram I watched get drawn • "Regenerate response" → "wait, why softmax?" → an answer Try it now at tryslate.live Three things we learned building it: 1. Typing kills curiosity. A blank text box turns "I wonder" into "I'll Google it later." Voice doesn't. 2. The whiteboard isn't a feature, it's the whole thing. Watching a concept get drawn lands different than reading the same words. 3. The system prompt that worked is one line: "don't monologue." The best tutor in the room is never the smartest one it's the most patient one. We shipped this end-to-end in less than 12 hours. Voice loop, live whiteboard, image gen, auth, session replay. On any other stack that's a six-week project. On Emergent it was a weekend. Pick a topic you've been meaning to understand. Tap the orb. You'll get it before your coffee goes cold. Built with emergent · #Vibecon

@Miss_populaire1 La plupart des casqniere trouvent mari sur les réseaux sociaux surtout entre Twitter, Snap, Instagram et Tik tok on connaît la méthode 😂😂

Nous les casaniere on va trouver mari Ou ?

@salioumbackefa1 Légui gueum force là Senegal metina na nak na ku nék gueum limu gueum rek wala du nonu amna si droit 😂

Ki bim nee gëmul julli sñ tuuba ci geej gi ba noppi ni amna xarit buy naaw la parewoon ci mom 😂😂

Build a Large Language Model from scratch! This repository contains the code examples for developing, pretraining, and finetuning a LLM from scratch. It is the official codebase for the book Build a Large Language Model (From Scratch). Notebook examples are included for each chapter: Chapter 1: Understanding Large Language Models Chapter 2: Working with Text Data Chapter 3: Coding Attention Mechanisms Chapter 4: Implementing a GPT Model from Scratch Chapter 5: Pretraining on Unlabeled Data Chapter 6: Finetuning for Text Classification Chapter 7: Finetuning to Follow Instructions Link to the repo in the comments!

We're living in the BEST era for doing research. 💪 After I graduated from my PhD, the rise of AI-native research gave me a new chance to revisit my research experience. Lately, doing research feels incredibly rewarding to me. I get to experience the pure joy of curiosity-driven science because I no longer have to worry about the lower-level implementations or getting bogged down by infrastructure 🚀 (I'll be sharing some of my own recent research driven by this very soon!) But today, let me introduce the New Orchestra 🎻. We wanted to ship a product that absorbs the friction and brings science back to the curiosity.

Every developer has an IDE. Researchers never had one. We don’t need another research agent, but a full stack workstation powered by AI. Today we're launching Orchestra @orch_research, the world's first Research IDE 🔬🧵

I revisited a paper on how to write a research paper well, and the message was blunt: Many rejections come from avoidable mistakes. Here are a few that matter most: 1️⃣ A weak title weakens the paper before it begins If the title is: → Too vague → Too long → Trying too hard you are already losing the reader. The title should match the study clearly. 2️⃣ Most people will judge your paper by the abstract That is just reality. The abstract is what reviewers see first. It is what shows up online. It is what people find when searching. If it is careless, too long, or poorly written, the rest of the paper may never get a fair chance. 3️⃣ Your introduction should not feel like punishment The paper makes an important point: You are not only reporting a study. You are telling a scientific story. If the introduction is dull, unfocused, or missing a clear hypothesis, you lose momentum early. 4️⃣ Methods are where credibility becomes visible This section must contain enough detail for someone to repeat your study if they want to. If your methods are unclear, the reader starts questioning everything else. 5️⃣ Results are not the place to show off Results should be results. Not discussion. Not interpretation. Not decoration. The paper is clear on this too: → avoid repeating yourself → keep figures useful → use as few images as necessary 6️⃣ The discussion is where many writers become careless This is where people often overstate findings. The paper warns against assigning greater significance to results than they deserve. That one mistake can make a serious paper feel unserious. 7️⃣ Good English is not a luxury It is part of the science being understood. One of the sharpest lessons in the paper is this: If a reviewer missed your point, the problem may not be the reviewer. It may be your writing. That line should humble every researcher. Because clarity is part of the work. ⸻ Good papers are built on: → clear titles → sharp abstracts → focused introductions → reproducible methods → honest discussions → clean writing A paper does not need to sound complicated to sound academic. That is what strong scientific writing looks like. 💬 What do you think damages a paper faster: a weak abstract, weak methods, or an overconfident discussion? ——— Source: Villar R. (2020). How to write that paper. Journal of Hip Preservation Surgery.

@ReineCalypso Merci du tips très insightful depuis le Sénégal

Je viens de découvrir, que j'ai meme pas besoin de VPS pour faire des petites app perso. Le combo Gemini (ou Claude) + Firebase (herbegement + DB) + AntiGravity c'est tellement banger et surtout... GRATOS ! (et la gestion de l'auth et tout c'est royal)

@SyFatimazahra @ElhadjiKandji1 Balma Madame la féministe 😂🙏🏾

@fallsalmone10 @ElhadjiKandji1 Koula wakh ni jabar dji yorouko wone en 2026 khamo len ni djiguen gni dakk lagniy demei wala khamoni nieupa indépendante wala la plus des femmes

Je suis tombé sur cette vidéo mais ça m'a fendu le cœur. Y'a des hommes c'est des chiens. Beaucoup de femmes sénégalaises vivent l'enfer dans leurs ménages par peur de divorcer et d'être juger par la société. Qu'Allah fasse que chaque femme tombe sur un homme bien.

@Rtz551 @ElhadjiKandji1 C'est dingue comment à chaque fois tu déformes mes propos et tu les sors de leur contexte vous êtes dangereuse madame. Bon vent à vous 😂😂

@fallsalmone10 @ElhadjiKandji1 « Lulen ci fek faut assumer » donc celles qui se font taper dans un ménage aisé dagn ko mériter ? Niemel lingay wakh seugn bi 🤷🏽‍♀️

@SyFatimazahra @ElhadjiKandji1 Mieux vaut prévenir que guérir revoyez vos critères avant de vous engager "MONEY" ne doit pas être le critère numéro un mais vous n'êtes pas prêt d'accepter cette vérité.

@fallsalmone10 @ElhadjiKandji1 Li ngay wakh ni ken ladiou lako dénoncer wo Dara li ngay dénoncer soko geumon le mots *mais * doussi meussa am yalla moy Natou diamam personne ne ce marie pou que l’on le maltraite nekal nitt

@Chitaaaa12 @ElhadjiKandji1 Zéro tolérance à la violence il y'a un adage qui dit que mieux vaut prévenir que guérir parfois les signaux sont là de vos prétendants mais vous choisissez selon des critères qui vous rattrapent tôt ou tard gni bax gnogui bax ba legui

@fallsalmone10 @ElhadjiKandji1 Votre commentaire me deçoit enormément genre vous trouver toujours des excuses…🤦🏾‍♀️💔