@noahdgoodman

humans& co-founder. professor of natural and artificial intelligence @Stanford. (@StanfordNLP @StanfordAILab) ex: Google DeepMind, UberAI

Joined November 2019
  • Tweets426
  • Following242
  • Followers7.4K
13Media
did i already mention how the @humansand team is cooking?
humans&
@humansandOct 23
We accelerated one of the most optimized NVFP4 decode kernels while also making it more accurate Fewer bits typically mean more speed but less accuracy. We did the opposite: more bits, but up to 10% faster generation Already used in Ampersand’s RL and Boba serving
1
31
2,008
first we built Persimmon, our sota human simulator. using that we trained Ampersand, a model that understands people individually and in groups. which powers Boba — released today! @humansand is cooking!
humans&
@humansandOct 22
Introducing Boba - the social intelligence that understands your goals and your community. Text Boba today or message it online at boba.app
6
7
85
3,987
noahdgoodman retweeted
Jeff Clune
@jeffcluneOct 21
Very excited to share what we have been working on at Recursive! A huge congrats to our entire team. We’re excited to hear what you all think. 🧪 🧬 🔭 🚀 📈✨
Richard Socher
@RichardSocherOct 21
Introducing Flint, @Recursive_SI’s first model: a fast local coding model built with our automated research system, the Eureka Machine. • Best-in-class fast local coding • Up to 2x faster than comparable models: 125+ tokens/sec on a MacBook Pro, 150+ on RTX Spark • Built from scratch in under 4 months Private, unmetered intelligence on your own machine. Open weights available now on @huggingface, run it locally on your favorite framework or try it for free on @opencode or @pidotdev.
6
5
87
5,673
while evaluating Ampersand v0.1 for sycophancy @__Charlie_G found something interesting: deployment makes models more sycophantic, but good human simulators substantially reduce evaluation awareness.🧵
1
3
1
53
3,091
more replies
but good user simulators in evaluations radically reduce eval awareness. the persimmon v0.1 user model with test-time scaling makes a huge difference. for Argon P(deploy) goes from 20% to 82% in our evals!
1
5
250
more details in the Ampersand v0.1 model card: ampersand.humansand.ai/blog/…
6
212
we've been developing a substantially different model training paradigm at @humansand, focussed on rich social simulation over long term outcomes. ampersand is our first model with this training, released today!
humans&
@humansandOct 20
No one changes the world alone Introducing Ampersand - the first model built for us. Ampersand is a collaborator trained to help people achieve their individual and collective goals. In multi-month simulations, users collaborating with Ampersand accomplished more than with Fable and other frontier models Available to everyone through our API
5
4
113
4,935
i'm proud of the excellent @humansand team, and how hard they've worked to pull this together. this is a first step and we've got Big Plans...
1
2
20
1,127
i've made some changes -- thanks for comments! here is the interactive version if you want to play with it: ngoodman.github.io/ai-safety…
noahdgoodman
@noahdgoodmanOct 18
i'm working on an AI safety lecture for my class, Minds and Machines. here is the first pass on a conceptual map... i would love suggestions from my safety friends.
2
2
47
3,615
i'll likely edit some more, so keep the comments coming.
2
420
the most satisfying tiny benchmark of matching the human distribution… like many benchmarks these days frontier models (persimmon) saturate
Alexis Ross
@alexisjrossOct 19
👀
1
1
1
33
2,663
i'm working on an AI safety lecture for my class, Minds and Machines. here is the first pass on a conceptual map... i would love suggestions from my safety friends.
12
17
1
155
14,855
@ancadianadragan @rohinmshah @AllanDafoe @sleepinyourhat @RogerGrosse @EthanJPerez @OwainEvans_UK (and many others, but i'm rushing)
2
2
1,257
i find this somewhat satisfying, but what i'm really waiting for is biting commentary by my friend @GaryMarcus. every time he attacks me (or hires me, we've done both) i know i'm on to something...
Brian(Yechan) Do
@doyc_1Oct 14
How to build a billion-dollar-valued AI research lab in 2026: 1. Share your critical take on huge companies like @humansand ($8.5B). 2. People who invested in them, or who attached their names to them (@Stanford, @OpenAI, @CarnegieMellon , etc.), will show up in the comments. 3. Congratulations. By being compared to them, your company is now also valued as a BILLION-DOLLAR COMPANY. 😂 Just joking. Btw, at @walnutinfra we’re about to release some pretty cool research soon. Follow us and stay tuned
1
1
31
8,512
RT @alexisjross: both of these tweets highlight the challenges of coming up with good evals for user models. 1) the ELIZA case study highli…

Introducing Persimmon | humans&

A model of human behavior, and new ways to measure how faithfully AI can simulate people.

persimmon.humansand.ai
This quoted post is unavailable.
Try viewing anyway · View on X
1
74
i find this especially amusing because i'm teaching ELIZA in Minds and Machines (Stanford's intro cogci w/ @mcxfrank) tomorrow...
Eric Zelikman
@ericzelikmanOct 6
People often eval user models on the "User Sim Index" - we probably shouldn't. I attached code for a "user model" scoring 95% on its 4 behavioral dims: 97% comms, 97% info, 92% clarify, 95% error reaction. I think this beats any released model. See also a sampled conversation
1
3
1
34
5,976
noahdgoodman retweeted
will brown
@willcbOct 5
i think the most exciting model release by a new western lab in recent months (behind jev, of course) is persimmon. people are massively underrating the utility of having great realistic user simulators. i would imagine they are quietly printing as a private API used in RL envs
19
34
2
808
49,398
Load more