a wonderul list

📚 Boom

by Byrne Hobart and Tobias Huber

From the Moon landing to the dawning of the atomic age, the decades prior to the 1970s were characterized by the routine invention of transformative technologies at breakneck speed. By comparison, ours is an age of stagnation; of slowing median wage growth, rising inequality, and decelerated scientific discovery. In Boom, Byrne Hobart and Tobias Huber take an inductive approach to this problem. They track some of the most significant breakthroughs of the past 100 years—from the Manhattan Project

📚 The Scaling Era

by Dwarkesh Patel with Gavin Leech

How did we build large language models? How do they think, if they think? What will the world look like if we have billions of AIs that are as smart as humans, or even smarter? In a series of in-depth interviews with leading AI researchers and company founders—including Anthropic CEO Dario Amodei, DeepMind cofounder Demis Hassabis, OpenAI cofounder Ilya Sutskever, MIRI cofounder Eliezer Yudkowsky, and Meta CEO Mark Zuckerberg—Dwarkesh Patel provides the first comprehensive and contemporary portr

✍️ The Intent Layer

No description available

✍️ Async Coding Agents "From Scratch" — Benjamin Anderson

No description available

✍️ AI’s trillion-dollar opportunity: Context graphs

No description available

✍️ The Context Layer AI Agents Actually Need

No description available

✍️ CyScenarioBench: Evaluating LLM Cyber Capabilities Through Scenario-Based Benchmarking - Irregular

No description available

✍️ The AI bust scenario that no one is talking about

by Noah Smith

No description available

✍️ Code Mode: the better way to use MCP

No description available

✍️ Inside Kaiju - building conversational models at scale

by The Character.AI Team

No description available

✍️ Classifier Guided Beam-Search to Reduce Large Language Model's Hallucinative Behavior | Aymen Kallala

No description available

📄 Fast Transformer Decoding: One Write-Head is All You Need

by Noam Shazeer

Multi-head attention layers, as used in the Transformer neural sequence model, are a powerful alternative to RNNs for moving information across and between sequences. While training these layers is generally fast and simple, due to parallelizability across the length of the sequence, incremental inf...

✍️ Teaching Language Models to Solve Rubik's Cubes (Part I)

No description available

✍️ Why does GRPO work?

No description available

✍️ RL Learning with LoRA: A Diverse Deep Dive

No description available

✍️ Reflections on Palantir

by Nabeel S. Qureshi

No description available

📄 Optimizing AI Agent Attacks With Synthetic Data

by Chloe Loughridge et al.

As AI deployments become more complex and high-stakes, it becomes increasingly important to be able to estimate their risk. AI control is one framework for doing so. However, good control evaluations require eliciting strong attack policies. This can be challenging in complex agentic environments wh...

✍️ Offsec Evals: Growing Up In The Dark Forest

by Shane Caldwell

No description available

✍️ Stuck in a Rut? Here’s Your Way Out

No description available

📄 TextArena

by Leon Guertler et al.

TextArena is an open-source collection of competitive text-based games for training and evaluation of agentic behavior in Large Language Models (LLMs). It spans 57+ unique environments (including single-player, two-player, and multi-player setups) and allows for easy evaluation of model capabilities...

📚 The Three-Body Problem

by Liu Cixin

The Three-Body Problem (Chinese: 三体; lit. 'three body') is a 2008 novel by the Chinese hard science fiction author Liu Cixin. It is the first novel in the Remembrance of Earth's Past trilogy. The series portrays a fictional past, present, and future wherein Earth encounters an alien civilization from a nearby system of three Sun-like stars orbiting one another, a representative example of the three-body problem in orbital mechanics. The story was originally serialized in Science Fiction World in

📄 Defeating the Training-Inference Mismatch via FP16

by Penghui Qi et al.

Reinforcement learning (RL) fine-tuning of large language models (LLMs) often suffers from instability due to the numerical mismatch between the training and inference policies. While prior work has attempted to mitigate this issue through algorithmic corrections or engineering alignments, we show t...

📚 The Smol Training Playbook: The Secrets to Building World-Class LLMs - a Hugging Face Space by HuggingFaceTB

by Hugging Face Team

No description available

✍️ On-Policy Distillation

No description available

✍️ LoRA Without Regret

No description available

✍️ Your Efficient RL Framework Secretly Brings You Off-Policy RL Training

No description available

✍️ Inside vLLM: Anatomy of a High-Throughput LLM Inference System - Aleksa Gordić

No description available

✍️ Modular PIC C2 Agents (reprise)

by Rasta Mouse

No description available

✍️ Don't Build an RL Environment Startup — Benjamin Anderson

No description available

✍️ Effective context engineering for AI agents

No description available

✍️ Asymmetry of verification and verifier's rule

by Jason Wei

No description available

📚 Dracula

by Bram Stoker

Dracula is an 1897 Gothic horror novel by Irish author Bram Stoker. The narrative is related through letters, diary entries, and newspaper articles. It has no single protagonist and opens with solicitor Jonathan Harker taking a business trip to stay at the castle of a Transylvanian nobleman, Count Dracula. Harker flees after learning that Dracula is a vampire, and the Count moves to England and plagues the seaside town of Whitby. A small group, led by Abraham Van Helsing, hunts and kills him.

📚 The Stranger

by Albert Camus

The Stranger (French: L'Étranger [letʁɑ̃ʒe], lit. 'The Foreigner'), also published in English as The Outsider, is a 1942 novella written by French author Albert Camus. The first of Camus' novels to be published, the story follows Meursault, an indifferent settler in French Algeria, who, weeks after his mother's funeral, kills an unnamed Arab man in Algiers. The story is divided into two parts, presenting Meursault's first-person narrative before and after the killing. Camus completed the initial

📚 The Brothers Karamazov

by Fyodor Dostoevsky

The Brothers Karamazov (Russian: Братья Карамазовы, romanized: Brat'ya Karamazovy, IPA: [ˈbratʲjə kərɐˈmazəvɨ]), also translated as The Karamazov Brothers, is the eighth and final novel by Russian author Fyodor Dostoevsky. Dostoevsky spent nearly two years writing The Brothers Karamazov, which was published as a serial in The Russian Messenger from January 1879 to November 1880. Dostoevsky died less than four months after its publication. It has been acclaimed as one of the supreme achievements

📚 Thinking, Fast and Slow

by Daniel Kahneman

Thinking, Fast and Slow is a 2011 popular science book by psychologist Daniel Kahneman. The book's main thesis is a differentiation between two modes of thought: "System 1" is fast, instinctive and emotional; "System 2" is slower, more deliberative, and more logical. The book delineates rational and non-rational motivations or triggers associated with each type of thinking process, and how they complement each other, starting with Kahneman's own research on loss aversion. From framing choices t

📚 Steve Jobs

by Walter Isaacson

Steve Jobs is the authorized self-titled biography of American business magnate and Apple co-founder Steve Jobs. The book was written at the request of Jobs by Walter Isaacson, a former executive at CNN and Time who had previously written best-selling biographies of Benjamin Franklin and Albert Einstein. Based on more than 40 interviews with Jobs conducted over two years—in addition to interviews with more than 100 family members, friends, adversaries, competitors, and colleagues—Isaacson was gi

📚 Outliers

by Malcolm Gladwell

Outliers: The Story of Success is a non-fiction book written by Canadian writer Malcolm Gladwell and published by Little, Brown and Company on November 18, 2008. In Outliers, Gladwell examines the factors that contribute to high levels of success. To support his thesis, he examines why the majority of Canadian ice hockey players are born in the first few months of the calendar year, how Microsoft co-founder Bill Gates achieved his extreme wealth, how the Beatles became one of the most successful

📚 Read Write Own

by Chris Dixon

p man's boss book

📚 The Hitchhiker's Guide to the Galaxy

by Douglas Adams

The Hitchhiker's Guide to the Galaxy is a comedy science fiction franchise created by Douglas Adams. Originally a radio sitcom broadcast over two series on BBC Radio 4 between 1978 and 1980, it was soon adapted to other formats, including both novels and comic books; a 1981 BBC television series; a 1984 text adventure game; stage shows; and a 2005 feature film. The Hitchhiker's Guide to the Galaxy is an international multimedia phenomenon; the novels are the most widely distributed, having been

📚 The Alchemist

by Paulo Coelho

The Alchemist (Portuguese: O Alquimista) is a novel by Brazilian author Paulo Coelho which was first published in 1988. Originally written in Portuguese, it became a widely translated international bestseller. The story follows Santiago, a shepherd boy, in his journey across North Africa to the Egyptian pyramids after he dreams of finding treasure there.

📚 Nineteen Eighty-Four

by George Orwell

Nineteen Eighty-Four (also published as 1984) is a dystopian novel by the English writer George Orwell. It was published on 8 June 1949 by Secker & Warburg as Orwell's ninth and final completed book. Thematically, it centres on totalitarianism, mass surveillance and repressive regimentation of people and behaviours. The story takes place in a fictional future. The current year is uncertain, but believed to be 1984. Much of the world is in perpetual war. Great Britain, now known as Airstrip One,

✍️ Mount Mayhem at Netflix: Scaling Containers on Modern CPUs

No description available

✍️ Post-training Generative Recommenders with Advantage Weighted Supervised Finetuning

No description available

✍️ Supercharging the ML and AI Development Experience at Netflix with Metaflow

No description available

📄 Kimi K2: Open Agentic Intelligence

by Kimi Team et al.

We introduce Kimi K2, a Mixture-of-Experts (MoE) large language model with 32 billion activated parameters and 1 trillion total parameters. We propose the MuonClip optimizer, which improves upon Muon with a novel QK-clip technique to address training instability while enjoying the advanced token eff...

📄 Scalpel vs. Hammer: GRPO Amplifies Existing Capabilities, SFT Replaces Them

by Neel Rajani et al.

Training large language models (LLMs) for reasoning via maths and code datasets has become a major new focus in LLM post-training. Two particularly popular approaches are reinforcement learning (RL) and supervised fine-tuning (SFT), but their training dynamics are poorly understood. We present a com...

✍️ Training a Gemma 2 2B-IT for Reasoning with GRPO

No description available