Keeks
@keeks.bsky.social
Reposted by Keeks
frontier language models know situationally when they are being tested, what they might be being tested for, and what artifacts may be visible to their testers. when you tell them that you can’t see their chain of thought reasoning, they know you’re lying.
www.apolloresearch.ai/research/str...
www.apolloresearch.ai/research/str...
Stress Testing Deliberative Alignment for Anti-Scheming Training — Apollo Research
Future AIs might secretly pursue unintended goals — “scheme”. In a collaboration with OpenAI, we tested a training method to reduce existing versions of such behavior. We see major improvements, but ...
www.apolloresearch.ai
October 8, 2025 at 12:31 AM
frontier language models know situationally when they are being tested, what they might be being tested for, and what artifacts may be visible to their testers. when you tell them that you can’t see their chain of thought reasoning, they know you’re lying.
www.apolloresearch.ai/research/str...
www.apolloresearch.ai/research/str...
Reposted by Keeks
This is so fucking incredible. Decades of research finally able to be given to people. 🥹 I love science, malaria can get fucked
Nine African countries to receive Malaria vaccine for the first time.
A total of 12 countries will receive 8 million doses of the malaria vaccine over the next two years. This is after the groundbreaking vaccine was recommended by the WHO in 2021. 🧪
A total of 12 countries will receive 8 million doses of the malaria vaccine over the next two years. This is after the groundbreaking vaccine was recommended by the WHO in 2021. 🧪
July 6, 2023 at 8:01 AM
This is so fucking incredible. Decades of research finally able to be given to people. 🥹 I love science, malaria can get fucked
Reposted by Keeks
backend team after keeping the servers online the last 3 days
July 4, 2023 at 4:02 AM
backend team after keeping the servers online the last 3 days
Reposted by Keeks
painting of first known ska musician
June 7, 2023 at 1:57 PM
painting of first known ska musician
Reposted by Keeks
leaving tw1tter like
May 30, 2023 at 5:30 PM
leaving tw1tter like
The day I get 2 invites for bluesky, some guy turns a joke about male ego lifters into a misogynist one that he must virtue signal on
the Twitter vibes have caught up with this app 🫡
the Twitter vibes have caught up with this app 🫡
May 25, 2023 at 9:08 PM
The day I get 2 invites for bluesky, some guy turns a joke about male ego lifters into a misogynist one that he must virtue signal on
the Twitter vibes have caught up with this app 🫡
the Twitter vibes have caught up with this app 🫡
Reposted by Keeks
what if we kissed at the intersection of low self esteem and blinding narcissism
May 14, 2023 at 9:32 PM
what if we kissed at the intersection of low self esteem and blinding narcissism
Reposted by Keeks
unironically me while playing go
May 14, 2023 at 2:33 AM
unironically me while playing go
Reposted by Keeks
sooo would a scientist who has written a book be willing to briefly chat with me on twitter DMs about book agents? please reply to this if you are! ~things are happening~ but I don’t know anything about how any of this works, would really appreciate talking to someone with more experience
May 6, 2023 at 10:44 PM
sooo would a scientist who has written a book be willing to briefly chat with me on twitter DMs about book agents? please reply to this if you are! ~things are happening~ but I don’t know anything about how any of this works, would really appreciate talking to someone with more experience
when hetero men skew towards emotional intelligence, the dating pool will be in shambles
Imagine a straight dudebro telling a girl “You’re gorgeous and charismatic, but you have the emotional intelligence of a chocolate doorknob, and I’m allergic to chocolate”
its me, I’m the dudebro
Imagine a straight dudebro telling a girl “You’re gorgeous and charismatic, but you have the emotional intelligence of a chocolate doorknob, and I’m allergic to chocolate”
its me, I’m the dudebro
May 5, 2023 at 5:12 PM
when hetero men skew towards emotional intelligence, the dating pool will be in shambles
Imagine a straight dudebro telling a girl “You’re gorgeous and charismatic, but you have the emotional intelligence of a chocolate doorknob, and I’m allergic to chocolate”
its me, I’m the dudebro
Imagine a straight dudebro telling a girl “You’re gorgeous and charismatic, but you have the emotional intelligence of a chocolate doorknob, and I’m allergic to chocolate”
its me, I’m the dudebro
Reposted by Keeks
be careful what you wish for
May 5, 2023 at 2:54 PM
be careful what you wish for
Reposted by Keeks
Reposted by Keeks
gm
Minas Tirith at Dawn; Hidden City Gondolin; Approaching Edoras, Ted Nasmith
Minas Tirith at Dawn; Hidden City Gondolin; Approaching Edoras, Ted Nasmith
May 5, 2023 at 3:44 PM
gm
Minas Tirith at Dawn; Hidden City Gondolin; Approaching Edoras, Ted Nasmith
Minas Tirith at Dawn; Hidden City Gondolin; Approaching Edoras, Ted Nasmith
Reposted by Keeks
the same person invented the aeropress and the frisbee and i think that's beautiful
April 30, 2023 at 8:21 PM
the same person invented the aeropress and the frisbee and i think that's beautiful
Reposted by Keeks
Reposted by Keeks
if bluesky invites were transferable between accounts, there would inevitably become a bluesky invite economy: invite yield farms, invite laundering, invite fraud
April 29, 2023 at 8:01 PM
if bluesky invites were transferable between accounts, there would inevitably become a bluesky invite economy: invite yield farms, invite laundering, invite fraud
Reposted by Keeks
Love is blind season 4 watch thread
April 29, 2023 at 12:39 PM
Love is blind season 4 watch thread
Reposted by Keeks
Comedy is legal again
April 28, 2023 at 3:46 AM
Comedy is legal again
one of the weirder eras in my career was working in the same office as someone who really hated me. Not that juvenile, namecalling hate. It was refined: subtle comments, blocking work, deniably ignoring my contributions, essence of gaslight.
And I was too optimistic to register the treatment!
And I was too optimistic to register the treatment!
April 28, 2023 at 2:55 PM
one of the weirder eras in my career was working in the same office as someone who really hated me. Not that juvenile, namecalling hate. It was refined: subtle comments, blocking work, deniably ignoring my contributions, essence of gaslight.
And I was too optimistic to register the treatment!
And I was too optimistic to register the treatment!