curius graph

all pages

showing 69101-69150 of 202637 pages (sorted by popularity)
Class Central • Find the best courses, wherever they exist.
Oscar Balcells Obeso on X: "An unexpected finding: when we train LoRA probes with minimal regularization, models become more epistemically cautious, sometimes acknowledging they've hallucinated immediately after doing so. We only train to predict binary hallucination labels from its own hidden states—no https://t.co/L1RiddP4Hj" / Twitter
Your Review: The Russo-Ukrainian War - Astral Codex Ten
*Kóryos - Wikipedia
Power from the Ground Up: Japan’s Land Bubble
Former Trump campaign manager registers to advocate for Israel
Unsong
Mottes and Baileys in AI discourse — LessWrong
Creditworthiness should not be for sale — LessWrong
(Fantasy) -> (Planning): A Core Mental Move For Agentic Humans? — LessWrong
Your Clone Wants to Kill You Because You Assumed Too Much — LessWrong
Do not hand off what you cannot pick up — LessWrong
Tomás Bjartur on X: "Ron's face pulled into a scowl. "If you don't like Quidditch, you don't have to make fun of it!" "If you can't criticise, you can't optimise. I'm suggesting how to improve the game. And it's very simple. Get rid of the Snitch." "They won't change the game just 'cause you say" / Twitter
Transluce on X: "Is your LM secretly an SAE? Most circuit-finding interpretability methods use learned features rather than raw activations, based on the belief that neurons do not cleanly decompose computation. In our new work, we show MLP neurons actually do support sparse, faithful circuits! https://t.co/lTBbUqoRlt" / Twitter
My AI Model Delta Compared To Yudkowsky — LessWrong
Scott Alexander on X: "@g_leech_ And before anyone asks, the next four are luck, divine favor, Chinese peptides, and sleep. After that it just goes in a loop between Chinese peptides and sleep forever." / Twitter
Quote by D.L. Moody: “Our greatest fear should not be of failure, but...”
No, Americans Don't Think Foreign Aid Is 26% of the Budget — LessWrong
No, Americans Don't Think Foreign Aid Is 26% of the Budget — LessWrong
Why does Eliezer make abrasive public comments? — LessWrong
An audio version of my blog post, Thoughts on AI progress (Dec 2025)
Recent LLMs can do 2-hop and 3-hop latent (no-CoT) reasoning on natural facts — LessWrong
The Best Advice I’ve Ever Heard for How to Be Happy - The New York Times
【教學】運用這數據帶你了解未來供需狀態! - 優分析UAnalyze
How might we safely pass the buck to AI? - by Josh Clymer
2403.10462
AI catastrophes and rogue deployments — AI Alignment Forum
glymour.pdf
- Sequent and resolution proof calculi, efficient CNF transformations
Reply to Sprenger's ``A Novel Solution to the Problem of Old Evidence''
2410.21514
Sabotage evaluations for frontier models \ Anthropic
Subversion Strategy Eval: Can language models statelessly strategize to subvert control protocols? — AI Alignment Forum
Schelling game evaluations for AI control — AI Alignment Forum
Claude 4 System Card
Impeccable agreements – The Manager's Handbook
Has AI caused the jobs recession? - UnHerd
The Deal of The Century - by Jason Hausenloy
Whose alignment research are we automating?
The Use of Knowledge in (AGI) Society
What Happens When AI Replaces Workers? | TIME
F-score - Wikipedia
roon on X: "had a dream where I was at a party, and every time a song I like comes on the dj immediately hits skip. wat means" / X
Mistral (wind) - Wikipedia
Granola on X: "@zapier Luke and Malaika made this quick demo earlier today 👇 https://t.co/36TABFX29S" / X
John Sherman on X: "If it's Sunday, it's Warning Shots! -Every Sunday at 10AM EST on The AI Risk Network on YouTube, only 15 minutes long. -I sit with @liron and @lethal_ai to discuss ONE missed AI risk warning shot each week on @AIRiskNetwork. https://t.co/z3ZiaCPVrY" / X
The existential horror of the paperclip factory. - YouTube
Why I don’t think AGI is right around the corner - YouTube
Why it's hard to make settings for high-stakes control research
What's worse, spies or schemers?