Braeden Hall
3 followers · 13 following · 867 views
on the atlas · 37
- Machine learning needs better tools – Replicate2 savers
- Chess Engines: A Zero to One Guide1 savers
- Grandmaster-Level Chess Without Search2 savers
- Playing chess with large language models1 savers
- [2310.20260] Learning to Play Chess from Textbooks (LEAP): a Corpus for Evaluating Chess Moves based on Sentiment Analysis1 savers
- [2209.11902] Learning Chess With Language Models and Transformers1 savers
- In Search of Organic Software13 savers
- 2401.10020.pdf1 savers
- David Hoang - Designer, investor, and writer14 savers
- Paul Buchheit: I am nothing3 savers
- neal.fun20 savers
- Rauno Freiberg18 savers
- Software 2.0. I sometimes see people refer to neural… | by Andrej Karpathy | Medium16 savers
- Replicate – Run open-source machine learning models with a cloud API6 savers
- Aim, fire, scan: the 80/20 of executing on big projects12 savers
- How to be More Agentic - by Cate Hall - Useful Fictions33 savers
- More Efficient In-Context Learning with GLaM – Google Research Blog1 savers
- Explorable Explanations10 savers
- Neural Networks1 savers
- Thirty Observations at Thirty62 savers
- Why to Start a Startup in a Bad Economy2 savers
- The Top Idea in Your Mind6 savers
- Why Startup Hubs Work1 savers
- Advice to Young People, The Lies I Tell Myself - jxnl.co19 savers
- GET OUT OF THE STUDIO1 savers
- An app can be a home-cooked meal27 savers
- towards flight: a new chapter - by Spencer Chang2 savers
- 2401.02415.pdf1 savers
- Language Model Composition1 savers
- Collaboration Cookbook1 savers
- Curius / Onboarding2704 savers
- Looking for Alice - by Henrik Karlsson - Escaping Flatland94 savers
- Childhoods of exceptional people - by Henrik Karlsson70 savers
- Toy Models of Superposition30 savers
- transformer_attention.pdf23 savers
- Manifesto · Moth Fund19 savers
- Visualizing A Neural Machine Translation Model (Mechanics of Seq2seq Models With Attention) – Jay Alammar – Visualizing machine learning one concept at a time.9 savers
highlights · 31
Will everyone appreciate the difference? Of course not, but the latter plants are labelled ‘organic’ to give us the information and the
In Search of Organic SoftwareIt turns out that a large portion of real-world problems have the property that it is significantly easier to collect the data (or more generally, identify a desirable behavior) than to explicitly write the program.
Software 2.0. I sometimes see people refer to neural… | by Andrej Karpathy | MediumIn Software 2.0 most often the source code comprises 1) the dataset that defines the desirable behavior and 2) the neural net architecture that gives the rough skeleton of the code, but with many details (the weights) to be filled in.
Software 2.0. I sometimes see people refer to neural… | by Andrej Karpathy | MediumThere is no success without a successor. Share your knowledge and wisdom with the future generation. If you’re successful, I believe it’s your obligation to pay it forward to the ecosystem that made you successful. Mentor someone, fund their work, etc.
David Hoang - Designer, investor, and writerBut I am nothing, and so I am finally free to be myself
Paul Buchheit: I am nothingYou should be able to import an audio transcriber the same way you an import an npm package. You should be able to fine-tune GPT as easily as you can fork something on GitHub.
Machine learning needs better tools – ReplicateAll you need is the ABZ: A = where are you now B = next small step Z = end vision
Aim, fire, scan: the 80/20 of executing on big projectsNearly all of my most fruitful collaborations over the last 3 years have come out of meetings I booked almost at random.
How to be More Agentic - by Cate Hall - Useful FictionsThe last couple times I was looking for a project, I made a point of meeting as many people doing related work as I could, even if there was no obvious benefit to doing so
How to be More Agentic - by Cate Hall - Useful FictionsIf you have not yet learned the skill of absorbing rejection, court it deliberately:
How to be More Agentic - by Cate Hall - Useful Fictionsfor a given set of people working on a given technology, the time to act is always now.
Why to Start a Startup in a Bad Economy"Okay, I'm focused on getting X, but let's not forget to read the headlines."
Advice to Young People, The Lies I Tell Myself - jxnl.coDepth is found in specificity; I’d rather make work that actually means something to someone than work that means nothing to everyone.
GET OUT OF THE STUDIOto design is to consider who is doing the looking, and whether or not they see what you see.
GET OUT OF THE STUDIOWhen you liberate programming from the requirement to be professional and scalable, it becomes a different activity altogether
An app can be a home-cooked mealI am the programming equivalent of a home cook.
An app can be a home-cooked mealfolk computing: home-cooked software, how to power a million tiny apps? small tools / folk interfaces and APIs / single-purpose websites and apps community infrastructure: community server garden, data co-ops, local infrastructure
towards flight: a new chapter - by Spencer ChangI’ve uncovered a treasure trove of questions, and now it’s time to begin to, if not answer, at least, sit with them.
towards flight: a new chapter - by Spencer ChangI feel compelled to explore, learn, and create in service of the internet and the kind of technology that I believe in
towards flight: a new chapter - by Spencer ChangTo this end, we propose CALM—Composition to Augment Language Models—which introduces cross-attention between models to compose their rep- resentations and enable new capabilities
Language Model CompositionIf you want to prompt someone to be authentic and playful and generative, you usually just need to ask them something where they have a rich experience to pull from but have never pulled an answer from that experience before.
Looking for Alice - by Henrik Karlsson - Escaping FlatlandThere is really no point in going to a café to talk safely (if you can avoid it). You want to rapidly extract as much information as possible, so you can figure out what you like and so that you can pattern match, and you want to communicate as much as possible, too, so you can filter people who wouldn’t fit you anyway
Looking for Alice - by Henrik Karlsson - Escaping FlatlandMy hypothesis is that if there are more small grants funding a diverse range of different projects, we will get more and better companies and non-profits in the long-run.
Manifesto · Moth FundAmbition means you're motivated to play games that others have already created in the world, while agency means you're driven to play a game of your own.
Manifesto · Moth FundFocus attracts “resources.”
Think more about what to focus on - by Henrik KarlssonAs a rule of thumb, you can only do 1 or 2 things well.
Think more about what to focus on - by Henrik KarlssonLet’s make more people lucky.
Childhoods of exceptional people - by Henrik KarlssonThe instinct was to curate a culture, not to teach, not primarily
Childhoods of exceptional people - by Henrik Karlssonthe neural networks we observe in practice are in some sense noisily simulating larger, highly sparse networks
Toy Models of SuperpositionWhen features are sparse, superposition allows compression beyond what a linear model would d
Toy Models of Superpositionhe’s created this immense vertical filing cabinet in his brain of layers and layers and layers of files
Curius / Onboarding