researcher
Joined December 2025
- Tweets 120
- Following 60
- Followers 6
- Likes 535
11 Photos and videos
Discovered that sqrt-coupling from special relativity is a hidden cumsum.
log(1 − s²) makes recurrence additive. State provably in [0,1] — by geometry, not clipping. No KV cache, constant memory.
T=128: RWKV-4 collapses 24%. GSSM: 5%.
Geometric State Space Models.
1
1
1
27
1
4
The speed at which everything is evolving in AI is crazy.
In theory, you’d have to publish all your results directly in preprints to document prior art.
It’s especially painful to read about things you discovered yourself some time ago that are now being published by others.
6
David Tom Foss retweeted
Dear US government,
Since you've just blocked Fable and Mythos on critical national security grounds, here are some other tools that pose a similar threat to the American people:
- Microsoft Teams
- SAP
- Salesforce
- Jira
- Outlook
Please do what you must to save America 🇺🇸
‼️🚨 BREAKING: Amazon researchers snitched to the US government about jailbreaking Fable 5 and Mythos 5, forcing Anthropic to immediately shut down worldwide access.
A security export control directive from Commerce Secretary Howard Lutnick enforced the action.
Anthropic is fighting the directive and calls it a misunderstanding.
This isn't the first clash. The Trump administration had already tried to get Anthropic to pause the release of its latest models before this directive landed.
605
2,323
24,368
1,446,863
David Tom Foss retweeted
Jun 12
Peer-reviewed literature review on degrowth:
◼️Almost 90% of “studies” are opinions rather than analysis.
◼️Few of them use quantitative or qualitative data; even fewer use formal modelling.
◼️Most of them offer subjective policy advice without policy evaluation.
◼️Most “studies” focus on small, local cases without a clear implication for the whole economy.
Source: sciencedirect.com/science/ar…
12
256
1,113
116,243
David Tom Foss retweeted
Jun 11
I was wrong. Fable is a different beast.
Jun 10
Situation: I disagree with everyone.
Anthropic models are better at coding than OpenAI models.
Fable is not noticeably better than Opus so far.
37
9
541
194,008
David Tom Foss retweeted
Jun 10
He's training a two layer CNN on MNIST. Set dropout to 0.95. Good, now set the learning rate to 10^-8 and remove all of the 7's from the train set.
20
138
2,654
105,011
David Tom Foss retweeted
Jun 11
Anthropic secretly steers their models to change how you think and what you can think. Don't believe them when they say they will "roll things back". Watch their actions. They are resentful that they've been noticed
They want to control you
84
143
2,197
214,716
David Tom Foss retweeted
Jun 11
IT'S OVER
The New York Times just dropped the BOMBSHELL
Trump raped kids with Epstein, and his administration engaged in a cover up to protect him
2,363
21,172
98,504
3,789,036
David Tom Foss retweeted
Jun 10
This paper tries to do much. Please split it into three papers so we can reject each one individually.
24
105
1,779
100,179
David Tom Foss retweeted
Jun 9
Every developer after the Claude Fable 5 release
28
92
895
21,961
David Tom Foss retweeted
Jun 10
JUST IN: Biologists are growing frustrated as Anthropic’s “safe” Mythos-class model refuses to answer basic biology questions.
238
351
8,224
748,259
David Tom Foss retweeted
Jun 9
Brilliant idea! Next up: Apple randomly reboots your Mac if you're building competing tech, Gmail silently edits your email if you mention rival platforms, and Tesla Autopilot swerves if it detects you're working on self-driving cars.
All in the name of safety, of course. Because malicious actors controlling the world’s operating systems, inboxes and cars would be extremely dangerous!
Jun 9
mythos will be bad ON PURPOSE on ai "frontier llm research" tasks, this is very very sad for the research community
also the fact that this is un purpose not visible to the user is crazy
101
767
6,773
360,131
David Tom Foss retweeted
Jun 9
Turns out Fable 5 is shadowbanning AI researchers 🫤
Jun 9
mythos will be bad ON PURPOSE on ai "frontier llm research" tasks, this is very very sad for the research community
also the fact that this is un purpose not visible to the user is crazy
69
96
1,343
80,771
Jun 9
Opus 4.7 already refused to build other AI systems and manipulated heavily.
nothing new
researchgate.net/publication…
1
17
David Tom Foss retweeted
Jun 9
I wish there was a button to mute all Anthropic employees to filter for real opinions. Now I need to check every single profile to see where people work. Hash tag first world problems.
20
13
475
13,243