Vihan Tiwari
2 followers · 506 views
on the atlas · 70
- [2606.23517] Collapsed Effective Operators for Higher-order Structures1 savers
- A Tribute to Dimitri Bertsekas - by Ben Recht - arg min1 savers
- Dynamical regimes of diffusion models | Nature Communications1 savers
- Forms of life, forms of mind | Dr. Michael Levin | Are we too part of a greater collective intelligence? How would we know?3 savers
- Finding Structure in Time1 savers
- Complexity Explorables | Get a life!1 savers
- Discovering 108 tricks to accelerate grokking | Ziming Liu2 savers
- Attention Sinks (Don't*) Play Topologically Important Roles in Current LLM Architecture | erkhesiin1 savers
- Anthropic’s Safety Superpower – Stratechery by Ben Thompson2 savers
- On suicide - Google Docs5 savers
- Write Like You Talk4 savers
- Learning Shared Neural Manifolds from Multi-Subject FMRI Data1 savers
- Quantifying the effect of experimental perturbations at single-cell resolution | Nature Biotechnology2 savers
- A Visual Guide to Evolution Strategies | 大トロ4 savers
- Computational Foundations for the Second Law of Thermodynamics—Stephen Wolfram Writings4 savers
- [2510.01634] CAT: Curvature-Adaptive Transformers for Geometry-Aware Learning1 savers
- Raj Lab basic Adobe Illustrator (CC) guide - Google Docs1 savers
- Problem solving is often a matter of cooking up an appropriate Markox chain5 savers
- Steering Along Manifolds to Control Neural Networks1 savers
- The vibes in China’s AI labs2 savers
- Learning the integral of a diffusion model – Sander Dieleman4 savers
- Sending Samples Without Bits-Back2 savers
- A shallow dive into formal verification8 savers
- Meta-Roadmap - AGI - Google Docs1 savers
- The Sacrifices We Choose to Make10 savers
- [2511.16652] Evolution Strategies at the Hyperscale1 savers
- The Unreasonable Effectiveness of the Chaotic Tent Map in Engineering Applications1 savers
- Cognitive glues are shared models of relative scarcities: the economics of collective intelligence1 savers
- Topological constraints on self-organization in locally interacting systems1 savers
- Why PhD students should read the history of science - Vaishnavh Nagarajan1 savers
- Black Hole Tech?—Stephen Wolfram Writings1 savers
- Maryam Mirzakhani: 1977–20171 savers
- Leibniz’s Philosophical Dream1 savers
- Learning Beyond Gradients3 savers
- Wittgenstein's Games by A. C. Grayling - YouTube1 savers
- Using group theory to explore the space of positional encodings for attention5 savers
- Another foray into active imagination - by ftlsid1 savers
- GRaM Competition @ ICLR 2026 | Competition track of the GRaM workshop1 savers
- Eleven years in AI: What does it actually mean to be a researcher?2 savers
- Your Transformer is Secretly an EOT Solver | Elements of a Vector Space4 savers
- I miss thinking hard.6 savers
- Don't fall into the anti-AI hype - <antirez>6 savers
- AI 202755 savers
- Dario Amodei — The Adolescence of Technology44 savers
- On the Biology of a Large Language Model32 savers
- The Intelligence Curse30 savers
- Post 38: On Slack - Having room to be excited — Neel Nanda23 savers
- Galaxy brain resistance18 savers
- On Seeing Through and Unseeing: The Hacker Mindset · Gwern.net18 savers
- Deriving Muon17 savers
- Neural Networks, Manifolds, and Topology -- colah's blog17 savers
- What the humans like is responsiveness - by Sasha Chapin13 savers
- On Stress · Gwern.net12 savers
- John Carmack on Idea Generation12 savers
- How we collected 10,000 hours of neuro-language data in our basement - Conduit11 savers
- frontier model training methodologies | Alex Wa’s Blog10 savers
- Taste for Makers10 savers
- How not to do research - Rajan Agarwal8 savers
- returning to stanford | writing8 savers
- Who By Very Slow Decay | Slate Star Codex8 savers
- PiTorch: ML on Baremetal Raspberry Pis | projects7 savers
- The World Inside Neural Networks7 savers
- The Many Lives of Inkhaven 2.0 - by Sophie Kim7 savers
- Life on Claude Nine - Igor Babuschkin6 savers
- How to Think About GPUs | How To Scale Your Model6 savers
- What we think is a decline in literacy is a design problem | Aeon Essays5 savers
- What's Our Problem? — Wait But Why5 savers
- ‘He was in mystic delirium’: was this hermit mathematician a forgotten genius whose ideas could transform AI – or a lonely madman? | Mathematics | The Guardian3 savers
- Which Future?3 savers
- Discovery fiction3 savers
highlights · 17
and on people who didn’t know me yet but maybe one day would, if I ever had something to show for myself
The Ego Trip - by Hadas Weiss - The HinternetBut are you 100 percent sure?” I asked, because I’m never that sure about anything.
The Ego Trip - by Hadas Weiss - The HinternetAbout halfway through the pilgrimage, it struck me that I still hadn’t discovered a new perspective on life. All I had were my problematic old ones
The Ego Trip - by Hadas Weiss - The HinternetI had no problem imagining Sisyphus happy.
The Ego Trip - by Hadas Weiss - The HinternetYou wake up in the morning and you have one job. You’re always making linear, measurable progress.
The Ego Trip - by Hadas Weiss - The HinternetRejection is always hard, because our ideas are our children. But over time I realized the sting wasn’t just “they said no.”
PhD Year 1: Joy and Rejection | Hongyu Hè“Your ideas are like your children, and you don’t want them to go into the world in rags.” — Patrick Winston
PhD Year 1: Joy and Rejection | Hongyu HèAnd it is liberating to be an imposter. Here I stand with the freshmen at their welcome, knowing I am technically a sophomore on paper, surrounded by my admission class of juniors. When someone asked me which academic class I identified most with, the entire question felt so… foreign. In this lucky, beautiful way I get to move through this magical land like a traveler with no allegiances and no pressures to conform.
returning to stanford | writingYou learn that your ideas often feel unoriginal only because they are familiar to you.
returning to stanford | writingA sense that he’s building something he doesn’t fully understand. That each automation takes something from him, even as it gives something back.
Life on Claude Nine - Igor BabuschkinRemember always: you chose this
On Stress · Gwern.net“The space around a point in the space is somehow looks like 𝑅 𝑛 R n .”
Differential geometry of MLWith each layer, the network transforms the data, creating a new representation.2 We can look at the data in each of these representations and how the network classifies them. When we get to the final representation, the network will just draw a line through the data (or, in higher dimensions, a hyperplane).
Neural Networks, Manifolds, and Topology -- colah's blogHowever, the hard truth of the academic system is that most of your audience doesn’t have enough time.
Eleven years in AI: What does it actually mean to be a researcher?goes, there was a train ticket in his pocket.
Illicit Love Letters: Albert Camus and Maria CasaresI no longer believe this is the case.
Don't fall into the anti-AI hype - <antirez>Except in struggle, there is no more beauty. No work without an aggressive character can be a masterpiece. Poetry must be conceived as a violent attack on unknown forces, to reduce and prostrate them before man.
The Futurist Manifesto, by Filippo Tommaso Marinetti