sean lee

2.5K posts

sean lee banner
sean lee

sean lee

@infinitefun_

synthetic libidinology | @websim_ai prev @southpkcommons @google

nyc Se unió Haziran 2014
5K Siguiendo1.2K Seguidores
sean lee
sean lee@infinitefun_·
@wayfaring_tim shhhh she's leaking alpha for counter-/co-manipulation
English
0
0
1
197
tim
tim@wayfaring_tim·
divya is great and i've been following her stuff for a long time but like i don't get why we've normalised this direction of manipulation
divya venn@divya_venn

how to play your man like a fiddle (bene gesserit maxxing): - use the carrot not the stick: @melissa wrote about this very eloquently, but don’t criticize behavior you don’t like, be happy and excited and grateful about the behavior you do. He’ll do more of it - use softness like a knife: never let yourself get nasally or shrill. softening and lowering and slowing your voice also makes people sit up and attention, often more so - use absence, not obsession: Have your own world he’s never fully privy to. give him room to miss you, *especially* when he’s not acting as he should - work around his pride, not against it never criticize or contradict him around other people. When you do criticize him, frame your criticisms in a way that align with his self image - be his mirror in the big ways never diminish yourself to make him feel strong. hold yourself to the same standards of excellence you always would. He’ll respect the ways in which you’re his equal - be his opposite in the small ones There’s a famous quote from mad men: you’ll never be one of the boys, so don’t even try. being a girl can take you very far if you do it right. show him his importance by letting him be of service in the ways you’re naturally built different. most obvious one is letting him lift things for you and reach tall shelves and do all the things he’s good at that aren’t your strengths. In return, take every opportunity in doing things for him that you’re good at. There’s nothing wrong w being feminine, stereotypically or otherwise. There doesn’t need to be any competition or shame or resentment in assuming complementary roles. In fact it’s a really sweet and beautiful thing - be non-reactive be excellent at regulating your own emotions. You don’t need him to do it for you. Your boundaries are so excellent that if he continually makes you upset you’ll just calmly leave, and he knows it. He has no power to make you lose control or act from fear or desperation

English
4
0
21
3.2K
tomie
tomie@tomieinlove·
Sorry to the language prescriptivists (lol) out there but “ABG” now refers to any Asian person, anywhere, ever, including the dudes
English
13
3
148
7.9K
Katie Chen
Katie Chen@dear_kxtie·
Why are women tearing each other apart this is actually so sad why can’t we all just be friends 🤷‍♀️ Only people lower than you will try to bring you down I swear
tiff@bytheophana

x.com/i/article/2049…

English
21
2
42
50.6K
sean lee
sean lee@infinitefun_·
@FoxNews this is gonna get area 51 stormed very soon
English
0
0
0
107
Fox News
Fox News@FoxNews·
The Southern California city of Westminster has officially installed "Charlie Kirk Way" signs to honor the late conservative activist.
English
1.8K
4.6K
65.6K
8.8M
Clavicular
Clavicular@Clavicular0·
Been productive and lively while sobermaxing, also getting leaner. We are ascending this summer.
English
458
403
10.1K
455K
Evan Yadegari
Evan Yadegari@EvanYadegari·
$100M app idea: Playground A dating app for minors One-shotted this with 10x.app
English
152
5
256
106.1K
sean lee
sean lee@infinitefun_·
@n0w00j this is like when u hold ur kid back a year so they mog the other kids at hockey
English
0
0
2
372
taoki
taoki@justalexoki·
we need a slur for people
English
276
13
618
19.2K
∿
@somewheresy·
NYC is prod. SF is staging. Furries & gooners & people on drugs spending lots of time on the computer is dev
English
2
1
88
3.1K
zach
zach@blip_tm·
new york is wild because you can be a regular at a bar and then it gets named the best in the country
English
10
32
1.9K
158.9K
sean lee
sean lee@infinitefun_·
re all the looped model hype rn -- reminder that we had a 7M parameter tiny looped model that easily beat LLM's on ARC AGI 1 and 2 first prize winner of the 2025 ARC Paper Award
Alexia Jolicoeur-Martineau@jm_alexia

New paper 📜: Tiny Recursion Model (TRM) is a recursive reasoning approach with a tiny 7M parameters neural network that obtains 45% on ARC-AGI-1 and 8% on ARC-AGI-2, beating most LLMs. Blog: alexiajm.github.io/2025/09/29/tin… Code: github.com/SamsungSAILMon… Paper: arxiv.org/abs/2510.04871

English
1
0
4
482
sean lee
sean lee@infinitefun_·
@gregisenberg nobody wants this (revealed vs stated preference)
English
0
0
2
69
GREG ISENBERG
GREG ISENBERG@gregisenberg·
It feels pretty obvious at this point that someone’s going to make billions building a social app that’s just for friends, no AI slop, no brainrot, calm design, chronological feed and no concept of followers
English
723
128
3.7K
542K
sean lee retuiteado
Sakana AI
Sakana AI@SakanaAILabs·
Discovering Novel LLM Experts via Task-Capability Coevolution Project: acdc-llm.github.io Paper: arxiv.org/abs/2604.14969 Can we build AI that is smarter than its parts? This week, our team will present AC/DC⚡ at #ICLR2026. The current paradigm in AI assumes that to solve more complex problems, we must train a single, ever-larger model. But no single model can excel at every task without massive computational costs. Instead of building one monolithic model, we asked what if we coevolved a diverse collective of specialized experts. We introduce Assessment Coevolving with Diverse Capabilities (AC/DC). It is a framework that simultaneously evolves a population of LLMs using evolutionary model merging and an archive of synthetic tasks generated by an AI scientist. As the tasks become more complex, the models must develop distinct, specialized skills to solve them. Crucially, AC/DC selects models based on Quality-Diversity. It keeps models not just because they score high on average, but because they solve different problems than the rest of the population. The results show that a collaborative task force of 8 small, evolved models can outperform a massive 72B parameter model, using significantly fewer total parameters. These models genuinely specialize, providing completely different, yet correct, approaches to complex problems. This suggests a new path forward for AI development, creating highly capable, parameter-efficient systems through collective intelligence rather than relying solely on brute-force scaling. OpenReview: openreview.net/forum?id=efNIN…
English
2
37
191
19.3K