LessWrong 2.0 Reader

View: New · Old · Top

Restrict date range: Today · This week · This month · Last three months · This year · All time

← previous page (newer posts) · next page (older posts) →

Minimal Motivation of Natural Latents
johnswentworth · 2024-10-14T22:51:58.125Z · comments (14)
[link] cancer rates after gene therapy
bhauth · 2024-10-16T15:32:53.949Z · comments (0)
How difficult is AI Alignment?
Sammy Martin (SDM) · 2024-09-13T15:47:10.799Z · comments (6)
Which LessWrong/Alignment topics would you like to be tutored in? [Poll]
Ruby · 2024-09-19T01:35:02.999Z · comments (12)
Motivation control
Joe Carlsmith (joekc) · 2024-10-30T17:15:50.881Z · comments (7)
MATS AI Safety Strategy Curriculum v2
DanielFilan · 2024-10-07T22:44:06.396Z · comments (6)
Australian AI Safety Forum 2024
Liam Carroll (liam-carroll) · 2024-09-27T00:40:11.451Z · comments (0)
Formalizing the Informal (event invite)
abramdemski · 2024-09-10T19:22:53.564Z · comments (0)
Time Efficient Resistance Training
romeostevensit · 2024-10-07T15:15:44.950Z · comments (8)
[link] [Paper Blogpost] When Your AIs Deceive You: Challenges with Partial Observability in RLHF
Leon Lang (leon-lang) · 2024-10-22T13:57:41.125Z · comments (0)
[link] [Paper] Programming Refusal with Conditional Activation Steering
Bruce W. Lee (bruce-lee) · 2024-09-11T20:57:08.714Z · comments (0)
[link] Point of Failure: Semiconductor-Grade Quartz
Annapurna (jorge-velez) · 2024-09-30T15:57:40.495Z · comments (8)
Reflections on the Metastrategies Workshop
gw · 2024-10-24T18:30:46.255Z · comments (5)
D&D Sci Coliseum: Arena of Data
aphyer · 2024-10-18T22:02:54.305Z · comments (23)
[link] IAPS: Mapping Technical Safety Research at AI Companies
Zach Stein-Perlman · 2024-10-24T20:30:41.159Z · comments (12)
Monthly Roundup #23: October 2024
Zvi · 2024-10-16T13:50:05.869Z · comments (12)
[link] An Interactive Shapley Value Explainer
James Stephen Brown (james-brown) · 2024-09-28T05:01:21.169Z · comments (9)
Startup Success Rates Are So Low Because the Rewards Are So Large
AppliedDivinityStudies (kohaku-none) · 2024-10-10T20:22:01.557Z · comments (6)
2025 Color Trends
sarahconstantin · 2024-10-07T21:20:03.962Z · comments (7)
instruction tuning and autoregressive distribution shift
nostalgebraist · 2024-09-05T16:53:41.497Z · comments (5)
[question] Implications of China's recession on AGI development?
Eric Neyman (UnexpectedValues) · 2024-09-28T01:12:36.443Z · answers+comments (3)
[Linkpost] Play with SAEs on Llama 3
Tom McGrath · 2024-09-25T22:35:44.824Z · comments (2)
Winners of the Essay competition on the Automation of Wisdom and Philosophy
AI Impacts (AI Imacts) · 2024-10-28T17:10:04.272Z · comments (3)
Bigger Livers?
sarahconstantin · 2024-11-08T21:50:09.814Z · comments (3)
Anthropic rewrote its RSP
Zach Stein-Perlman · 2024-10-15T14:25:12.518Z · comments (19)
Are we dropping the ball on Recommendation AIs?
Charbel-Raphaël (charbel-raphael-segerie) · 2024-10-23T17:48:00.000Z · comments (14)
Book Review: On the Edge: The Business
Zvi · 2024-09-25T12:20:06.230Z · comments (0)
[Intuitive self-models] 7. Hearing Voices, and Other Hallucinations
Steven Byrnes (steve2152) · 2024-10-29T13:36:16.325Z · comments (2)
Metastatic Cancer Treatment Since 2010: The Success Stories
sarahconstantin · 2024-11-04T22:50:09.386Z · comments (0)
[link] Characterizing stable regions in the residual stream of LLMs
Jett Janiak (jett) · 2024-09-26T13:44:58.792Z · comments (4)
Compelling Villains and Coherent Values
Cole Wyeth (Amyr) · 2024-10-06T19:53:47.891Z · comments (4)
0.202 Bits of Evidence In Favor of Futarchy
niplav · 2024-09-29T21:57:59.896Z · comments (0)
[link] An X-Ray is Worth 15 Features: Sparse Autoencoders for Interpretable Radiology Report Generation
hugofry · 2024-10-07T08:53:14.658Z · comments (0)
[link] Generative ML in chemistry is bottlenecked by synthesis
Abhishaike Mahajan (abhishaike-mahajan) · 2024-09-16T16:31:34.801Z · comments (2)
[link] AISafety.info: What is the "natural abstractions hypothesis"?
Algon · 2024-10-05T12:31:14.195Z · comments (2)
Signaling with Small Orange Diamonds
jefftk (jkaufman) · 2024-11-07T20:20:08.026Z · comments (1)
LASR Labs Spring 2025 applications are open!
Erin Robertson · 2024-10-04T13:44:20.524Z · comments (0)
The murderous shortcut: a toy model of instrumental convergence
Thomas Kwa (thomas-kwa) · 2024-10-02T06:48:06.787Z · comments (0)
I'm creating a deep dive podcast episode about the original Leverage Research - would you like to take part?
spencerg · 2024-09-22T14:03:22.164Z · comments (2)
Distinguish worst-case analysis from instrumental training-gaming
Olli Järviniemi (jarviniemi) · 2024-09-05T19:13:34.443Z · comments (0)
AI Safety Camp 10
Robert Kralisch (nonmali-1) · 2024-10-26T11:08:09.887Z · comments (7)
Glitch Token Catalog - (Almost) a Full Clear
Lao Mein (derpherpize) · 2024-09-21T12:22:16.403Z · comments (3)
COT Scaling implies slower takeoff speeds
Logan Zoellner (logan-zoellner) · 2024-09-28T16:20:00.320Z · comments (56)
A New Class of Glitch Tokens - BPE Subtoken Artifacts (BSA)
Lao Mein (derpherpize) · 2024-09-20T13:13:26.181Z · comments (7)
Open Source Replication of Anthropic’s Crosscoder paper for model-diffing
Connor Kissane (ckkissane) · 2024-10-27T18:46:21.316Z · comments (1)
OODA your OODA Loop
Raemon · 2024-10-11T00:50:48.119Z · comments (3)
Exploring SAE features in LLMs with definition trees and token lists
mwatkins · 2024-10-04T22:15:28.108Z · comments (5)
[link] A Percentage Model of a Person
Sable · 2024-10-12T17:55:07.560Z · comments (3)
Eye contact is effortless when you’re no longer emotionally blocked on it
Chipmonk · 2024-09-27T21:47:01.970Z · comments (24)
[link] Big tech transitions are slow (with implications for AI)
jasoncrawford · 2024-10-24T14:25:06.873Z · comments (16)
← previous page (newer posts) · next page (older posts) →