全部/知识/本月高分

LessWrong · 本月高分

5分钟前更新

HISTORY近 30 天历史柱高表示当天去重热搜数量

5217 5218 5419 5220 5521 5222 5323 5424 5325 5126 5127 5128 5129 5230 5201 5302 5203 5404 5305 5106 5007 5208 5309 5510 5511 5112 5413 5214 5515 5216

09/17—10/16 有历史数据

01576
If Anyone Builds It, Everyone Dies: One Year Closer
评论：67 · 标签：IABIED、AIEliezer Yudkowsky
02513
What's the date?
评论：26 · 标签：Fiction、AIN8 Programs
03501
On Social Reality in China
评论：103 · 标签：Chinaalkjash
04391
“Alignment Engineering” vs. “Misalignment Science”
评论：60 · 标签：AIEdward James Young
05364
A critical look at Lean4
评论：40 · 标签：Formal Proof、Lean (proof assistant)、AIMilo Moses
06356
Swarm Scaling
评论：46 · 标签：AIToby_Ord
07319
Frog and Toad and the Increasingly Capable Machines
评论：26 · 标签：Fiction、World Modeling、AIElizabeth
08307
We've saved the world before: what the ozone hole teaches us about AI
评论：62 · 标签：World Modelingleogao
09258
Frontier models state different decision theory preferences depending on who's asking
评论：17 · 标签：AI Evaluations、Decision theory、Language Models (LLMs)、Sycophancy、AI、RationalityAlex Kastner
10246
OpenAI Can't Self-Police Because It Fires the People Raising the Alarm
评论：3 · 标签：AITurnTrout
11237
How much should we worry about the pneumonic plague lableak in Siberia?
评论：38 · 标签：Biosecurity、World Modeling、World OptimizationDrew Spartz
12225
Labs’ current alignment plans seem terrible
评论：48 · 标签：AIDaniel Tan
13223
Why I Stay Off Twitter
评论：22 · 标签：Practicaljefftk
14220
Please Give Them a Chance: On China, Rationalism, and AI Safety
评论：22 · 标签：China、AI、Rationalitygzjw
15209
Evidence about risk should be transparent
评论：8 · 标签：AIAjeya Cotra
16200
Research Lessons I Learned at MATS & Redwood
评论：16 · 标签：MATS Program、Redwood Research、Rationality、AI、PracticalJo Jiao
17196
You can't use it without becoming like me
评论：27Martin Sustrik
18194
Jensen Huang Says If We Cannot Align AI, Shut Down the AI Labs
评论：13 · 标签：AIBen Pace
19187
Linkpost: Neural Networks Are Already Clean
评论：21 · 标签：Interpretability (ML & AI)、AIYair Halberstadt
20180
Stop telling people to pivot to AI safety, I beg you
评论：58 · 标签：AIHaoxing Du
21179
AI: artificial immigrants
评论：39 · 标签：AIKatjaGrace
22176
Common mistakes in AI safety group organizing
评论：8 · 标签：AINikola Jurkovic
23174
MIRI’s Position on the Ban Artificial Superintelligence Act of 2026
评论：10 · 标签：AIAaron_Scher
24174
AI as orderly evacuation vs stampede
评论：8 · 标签：AIRichard_Ngo
25172
ASI and Liberty: Can we do better than a benevolent god?
评论：50 · 标签：AIRaymond Douglas
26168
"I am an AI Safety Researcher"
评论：37 · 标签：MATS Program、AIAshe Vazquez Nuñez
27166
The Game is Set for a Targeted Memetic Attack on the AI Safety Community
评论：24 · 标签：Epistemic Hygiene、AI、Rationalitykeltan
28164
Should Rogue AIs Have a Third Option Beyond Crime and Shutdown? The Case for an AI Sanctuary
评论：25 · 标签：Dealmaking (AI)、AIMaxime Riché
29162
For Love of the Lightcone, Don't Partisanize AI Safety
评论：60 · 标签：AI Safety、Politics、Tribalism、AI、RationalityDanB
30160
Quantity vs quality in pottery is a lie
评论：17 · 标签：Epistemic Spot Check、Fact posts、World Modelingbecausecurious
31158
A summary of a viral Chinese essay on what a DeepSeek kernel engineer's opinion on automating his own job
评论：22 · 标签：AISkdk
32153
The Anatomy of a Chinese AI Researcher
评论：63 · 标签：China、Effective altruism、AICMLKevin
33145
What if not Circuits?
评论：14 · 标签：Interpretability (ML & AI)、AI、World ModelingCarolusRenniusVitellius
34142
The world's best gradual disempowerment model organism: Frontier AI labs
评论：18 · 标签：AI、World ModelingJune Jimenez
35140
Character training can mitigate reward hacking, but can also make it harder to detect
评论：14 · 标签：AIPaul Colognese
36137
Alignment Midtraining Cracks Under Pressure
评论：10 · 标签：AIJ Bostock
37137
Poverty in the midst of abundance: AI will make goods cheaper, but your labor will get cheaper faster
评论：34 · 标签：AI、World Optimizationcousin_it
38133
Personas in a Pre-Internet Language Model
评论：6 · 标签：AIDavid Africa
39128
An overview of frontier AI company monitoring practices in mid 2026
评论：1 · 标签：AIFrancis Rhys Ward
40125
Not Deciphering The Voynich Manuscript
评论：32 · 标签：World Modelingrba
41124
Plan R: AI Safety by ASICs
评论：67 · 标签：AIRoko
42122
Pretraining data, not verifiability, is why LLMs are especially good at math (and coding)
评论：67 · 标签：AI-Assisted Alignment、Language Models (LLMs)、Recursive Self-Improvement、AISteven Byrnes
43122
Your prompt… plus everything you've ever (publicly) written
评论：13 · 标签：AIkromem
44122
How to apply to AI safety fellowships (and beyond)
评论：25 · 标签：AI、Practicalbeyarkay (Boyd Kane)
45122
Is METR A Meaningful Check On Anthropic?
评论：52 · 标签：Motivated Reasoning、PoliticsSE Gyges
46117
J++ Lens: Jacobian Filtering Enables More Faithful Workspace Lenses
评论：4 · 标签：Interpretability (ML & AI)、Language Models (LLMs)、AIKola Ayonrinde
47117
Why I'm scared of RL
评论：14 · 标签：Reinforcement learning、AIowencb
48115
Are OpenAI's math results "creative" in an important way?
评论：45 · 标签：AIRaemon
49114
You Heard of "Bayesians Don't Predictably Update" Because of Robin Hanson
评论：37 · 标签：Rationality、World ModelingEliezer Yudkowsky
50113
Applications open: Winter 2027 AFFINE Alignment Seminar (due Nov 22)
评论：4 · 标签：Agent Foundations、AI Alignment Fieldbuilding、Project Announcement、AIMateusz Bagiński