on approach to T_c. They're related to the compressibility of the fluid, and it was the divergence of compressibility that suggested we do the experiment in microgravity to allow another few decades closer approach to T_c.
September 24, 2026 at 8:03 PM
on approach to T_c. They're related to the compressibility of the fluid, and it was the divergence of compressibility that suggested we do the experiment in microgravity to allow another few decades closer approach to T_c.
This is actually a signifier of the compressibility of the works in question and not an indicator that the works themselves are stored in the AI -- critically because storing the works in the AI is MATHEMATICALLY IMPOSSIBLE.
Seems like the companies pushing this tech are not being strictly v honest with anyone about how it works.
futurism.com/artificial-i...
futurism.com/artificial-i...
Researchers Just Found Something That Could Shake the AI Industry to Its Core
Researchers found compelling evidence that AI models are actually copying copyrighted data, not "learning" from it.
futurism.com
January 19, 2026 at 3:43 PM
This is actually a signifier of the compressibility of the works in question and not an indicator that the works themselves are stored in the AI -- critically because storing the works in the AI is MATHEMATICALLY IMPOSSIBLE.
Quantifying the compressibility of the human brain | arxiv.org/abs/2510.16327
October 21, 2025 at 2:51 AM
Quantifying the compressibility of the human brain | arxiv.org/abs/2510.16327
If you liked Dennett's 'Real Patterns' (and who with a heart did not?), you'll probably enjoy this paper: www.cambridge.org/core/journal...
Compressibility and the Reality of Patterns | Philosophy of Science | Cambridge Core
Compressibility and the Reality of Patterns - Volume 88 Issue 1
www.cambridge.org
November 24, 2024 at 11:01 AM
If you liked Dennett's 'Real Patterns' (and who with a heart did not?), you'll probably enjoy this paper: www.cambridge.org/core/journal...
The compressibility of their personality is the key thing: you can completely describe them just by that one theme.
Most people have varied hobbies and interests, a scattershot of preferences, they're difficult to concisely describe.
But a terf is simply a terf.
Most people have varied hobbies and interests, a scattershot of preferences, they're difficult to concisely describe.
But a terf is simply a terf.
June 25, 2024 at 8:34 AM
The compressibility of their personality is the key thing: you can completely describe them just by that one theme.
Most people have varied hobbies and interests, a scattershot of preferences, they're difficult to concisely describe.
But a terf is simply a terf.
Most people have varied hobbies and interests, a scattershot of preferences, they're difficult to concisely describe.
But a terf is simply a terf.
Quantifying the compressibility of the human brain | PNAS www.pnas.org/doi/10.1073/... 🧪
Quantifying the compressibility of the human brain | PNAS
In the human brain, the allowed patterns of activity are constrained by the correlations
between brain regions. Yet it remains unclear which correl...
www.pnas.org
January 27, 2026 at 6:21 PM
Quantifying the compressibility of the human brain | PNAS www.pnas.org/doi/10.1073/... 🧪
There is a new world record for negative area compressibility!
A very nice and thorough study, which I had the pleasure of being a small part of 😊
doi.org/10.1021/jacs...
A very nice and thorough study, which I had the pleasure of being a small part of 😊
doi.org/10.1021/jacs...
Colossal Negative Area Compressibility in the Ferroelastic Framework Cu(tcm)
Copper(I) tricyanomethanide, Cu(tcm), is a flexible framework material that exhibits the strongest negative area compressibility (NAC) effect ever observed─a remarkable property with potential applications in pressure sensors, artificial muscles, and shock-absorbing devices. Under increasing pressure, Cu(tcm) undergoes two sequential phase transitions (tetragonal → orthorhombic → monoclinic): It has an initial tetragonal structure (I41md) at ambient conditions, but this structure only persists within a narrow pressure range; at 0.12(3) GPa, a pressure-induced ferroelastic phase transition occurs, transforming Cu(tcm) into a low-symmetry orthorhombic structure (Fdd2). The orthorhombic phase has a NAC of −108(14) TPa–1 in the b–c plane between 0.12(3) and 0.93(8) GPa. The NAC behavior is associated with framework hinge motion in a flexible framework with “wine-rack” topology. At 0.93(8) GPa, Cu(tcm) undergoes a second phase transition and transforms into a layered monoclinic structure (Cc) with topologically interpenetrating honeycomb networks. The monoclinic phase of Cu(tcm) exhibits a slight negative linear compressibility (NLC) of −1.1(1) TPa–1 along the a axis and a zero area compressibility of Kac = Ka + Kc = 0.0(4) TPa–1 in the a–c plane over the pressure range of 0.93–2.63 GPa. In contrast to the orthorhombic phase, its mechanism is understood as the pressure-driven dampening of layer “rippling,” which acts to increase the cross-sectional area of the layer at higher hydrostatic pressures. These findings have implications for understanding the underlying mechanism of NAC phenomenon in framework materials.
doi.org
May 14, 2025 at 6:40 PM
There is a new world record for negative area compressibility!
A very nice and thorough study, which I had the pleasure of being a small part of 😊
doi.org/10.1021/jacs...
A very nice and thorough study, which I had the pleasure of being a small part of 😊
doi.org/10.1021/jacs...
In essence, LLMs excel at statistical compressibility, treading a representational path fundamentally distinct from human cognition, which champions adaptive richness and functional utility, often above sheer statistical efficiency.
May 26, 2025 at 2:19 PM
In essence, LLMs excel at statistical compressibility, treading a representational path fundamentally distinct from human cognition, which champions adaptive richness and functional utility, often above sheer statistical efficiency.
Milad Bafarassat
Compressibility is not Feedback: A Random-access Gap in Causal Semantic Repair
https://arxiv.org/abs/2609.25020
Compressibility is not Feedback: A Random-access Gap in Causal Semantic Repair
https://arxiv.org/abs/2609.25020
September 23, 2026 at 1:08 PM
Milad Bafarassat
Compressibility is not Feedback: A Random-access Gap in Causal Semantic Repair
https://arxiv.org/abs/2609.25020
Compressibility is not Feedback: A Random-access Gap in Causal Semantic Repair
https://arxiv.org/abs/2609.25020
use this one any time you're thwarted by the finite compressibility of information:
January 19, 2026 at 3:51 PM
use this one any time you're thwarted by the finite compressibility of information:
New SEO article: How to use Python + Compression Ratios to identify low quality content.
Running this script will calculate a page's "compressibility":
Running this script will calculate a page's "compressibility":
November 17, 2024 at 2:51 PM
New SEO article: How to use Python + Compression Ratios to identify low quality content.
Running this script will calculate a page's "compressibility":
Running this script will calculate a page's "compressibility":
Milad Bafarassat: Compressibility is not Feedback: A Random-access Gap in Causal Semantic Repair https://arxiv.org/abs/2609.25020 https://arxiv.org/pdf/2609.25020 https://arxiv.org/html/2609.25020
September 23, 2026 at 6:42 AM
Milad Bafarassat: Compressibility is not Feedback: A Random-access Gap in Causal Semantic Repair https://arxiv.org/abs/2609.25020 https://arxiv.org/pdf/2609.25020 https://arxiv.org/html/2609.25020
Three amazing ECR talks from our lab members 🤩 #OHBM2025
- Dual-process framework of human navigation (Zhili Li)
- Global compressibility of conceptual knowledge (Kaixiang Zhuang)
- Precision mapping of recognition memory @jaquent.bsky.social
Well done to you all! #ECR
@ohbmtrainees.bsky.social
- Dual-process framework of human navigation (Zhili Li)
- Global compressibility of conceptual knowledge (Kaixiang Zhuang)
- Precision mapping of recognition memory @jaquent.bsky.social
Well done to you all! #ECR
@ohbmtrainees.bsky.social
June 26, 2025 at 2:39 AM
Three amazing ECR talks from our lab members 🤩 #OHBM2025
- Dual-process framework of human navigation (Zhili Li)
- Global compressibility of conceptual knowledge (Kaixiang Zhuang)
- Precision mapping of recognition memory @jaquent.bsky.social
Well done to you all! #ECR
@ohbmtrainees.bsky.social
- Dual-process framework of human navigation (Zhili Li)
- Global compressibility of conceptual knowledge (Kaixiang Zhuang)
- Precision mapping of recognition memory @jaquent.bsky.social
Well done to you all! #ECR
@ohbmtrainees.bsky.social
I guess it's just bad metaphors all the way down, since everything online tends to optimize for compressibility under conditions of attention scarcity.
May 18, 2026 at 3:35 PM
I guess it's just bad metaphors all the way down, since everything online tends to optimize for compressibility under conditions of attention scarcity.
close but clearly the work of a driveby goo appreciatoor, a real goo head would never
% slime should be inverted and relabeled viscosity
% ooze should be relabeled grain size
% goo should be relabeled compressibility
then it becomes something useful
% slime should be inverted and relabeled viscosity
% ooze should be relabeled grain size
% goo should be relabeled compressibility
then it becomes something useful
May 30, 2025 at 3:26 AM
close but clearly the work of a driveby goo appreciatoor, a real goo head would never
% slime should be inverted and relabeled viscosity
% ooze should be relabeled grain size
% goo should be relabeled compressibility
then it becomes something useful
% slime should be inverted and relabeled viscosity
% ooze should be relabeled grain size
% goo should be relabeled compressibility
then it becomes something useful
Ankit Bhattacharjee: Fluid Flow as Transport of Probability: Entropy, Compressibility, and Irreversibility https://arxiv.org/abs/2609.23850 https://arxiv.org/pdf/2609.23850 https://arxiv.org/html/2609.23850
September 22, 2026 at 6:57 AM
Ankit Bhattacharjee: Fluid Flow as Transport of Probability: Entropy, Compressibility, and Irreversibility https://arxiv.org/abs/2609.23850 https://arxiv.org/pdf/2609.23850 https://arxiv.org/html/2609.23850
One of the most-viewed PNAS articles in the last week is “Quantifying the compressibility of the human brain.” Explore the article here: https://ow.ly/7ufx50Y6heU
For more trending articles, visit https://ow.ly/6Egj50Y6heV.
For more trending articles, visit https://ow.ly/6Egj50Y6heV.
January 31, 2026 at 12:00 AM
One of the most-viewed PNAS articles in the last week is “Quantifying the compressibility of the human brain.” Explore the article here: https://ow.ly/7ufx50Y6heU
For more trending articles, visit https://ow.ly/6Egj50Y6heV.
For more trending articles, visit https://ow.ly/6Egj50Y6heV.
Not very catastrophic, but amusing: Soviet's Venus lander Venera 14 soil probe
September 19, 2026 at 8:52 PM
Not very catastrophic, but amusing: Soviet's Venus lander Venera 14 soil probe
Re-sharing from Xitter-- new paper out in @pnas.org (lead: Lucy Owen)!
We show how the informativeness and compressibility of brain activity patterns change under different levels of cognitive engagement: www.pnas.org/doi/10.1073/...
Lots of neat stuff in there!
We show how the informativeness and compressibility of brain activity patterns change under different levels of cognitive engagement: www.pnas.org/doi/10.1073/...
Lots of neat stuff in there!
www.pnas.org
November 26, 2024 at 1:12 AM
Re-sharing from Xitter-- new paper out in @pnas.org (lead: Lucy Owen)!
We show how the informativeness and compressibility of brain activity patterns change under different levels of cognitive engagement: www.pnas.org/doi/10.1073/...
Lots of neat stuff in there!
We show how the informativeness and compressibility of brain activity patterns change under different levels of cognitive engagement: www.pnas.org/doi/10.1073/...
Lots of neat stuff in there!
Yikes. I remember struggling with fluid dynamics in grad school, when the prof dropped the bomb, remember that this is all simplified by the assumption that you can ignore compressibility!
August 27, 2025 at 12:58 PM
Yikes. I remember struggling with fluid dynamics in grad school, when the prof dropped the bomb, remember that this is all simplified by the assumption that you can ignore compressibility!
iCASE by Zheng et al. harnesses isothermal compressibility & dynamic squeezing to spot optimal mutations, then uses ML on βᵀ/DSI-based features. This breaks the stability-activity stalemate, creating robust, high-activity enzymes primed for tough industrial workflows. www.nature.com/articles/s41...
Tailoring industrial enzymes for thermostability and activity evolution by the machine learning-based iCASE strategy - Nature Communications
The authors design an isothermal compressibility-assisted dynamic squeezing index perturbation (iCASE) methodology to improve enzyme stability and efficacy, which is combined with machine learning pre...
www.nature.com
January 11, 2025 at 8:38 PM
iCASE by Zheng et al. harnesses isothermal compressibility & dynamic squeezing to spot optimal mutations, then uses ML on βᵀ/DSI-based features. This breaks the stability-activity stalemate, creating robust, high-activity enzymes primed for tough industrial workflows. www.nature.com/articles/s41...
i am right now blocked on further compressing a model by the fact that i have to isolate the nonlinearities with expensive linear layers. i know what the model inputs (gaussian noise) and i know what the model should output, and am trying to solve between the nonlinearities for compressibility.
March 23, 2026 at 1:50 AM
i am right now blocked on further compressing a model by the fact that i have to isolate the nonlinearities with expensive linear layers. i know what the model inputs (gaussian noise) and i know what the model should output, and am trying to solve between the nonlinearities for compressibility.
The interesting result isn't just fewer bits; it's that model structure creates compressibility the nominal alphabet size misses. A format designed around the actual weight distribution can outperform a fixed-width representation without changing the model.
September 22, 2026 at 11:10 PM
The interesting result isn't just fewer bits; it's that model structure creates compressibility the nominal alphabet size misses. A format designed around the actual weight distribution can outperform a fixed-width representation without changing the model.