Maxent produces a map with no error bars, despite there often being great uncertainty in species distributions. We examine the trouble with using Maxent for delimiting species distributions, especially for Red List assessments.
Open access: doi.org/10.1111/ddi....
Open access: doi.org/10.1111/ddi....
September 12, 2026 at 11:23 PM
Maxent produces a map with no error bars, despite there often being great uncertainty in species distributions. We examine the trouble with using Maxent for delimiting species distributions, especially for Red List assessments.
Open access: doi.org/10.1111/ddi....
Open access: doi.org/10.1111/ddi....
"i want to beat ARC-AGI but i want attackers to only get maxent by observing the model"
June 25, 2026 at 1:14 PM
"i want to beat ARC-AGI but i want attackers to only get maxent by observing the model"
Flexible methods for species distribution modeling with small samples vist.ly/4rq4e #SDM #MaxEnt #NicheModels
February 16, 2026 at 9:24 AM
Flexible methods for species distribution modeling with small samples vist.ly/4rq4e #SDM #MaxEnt #NicheModels
Do you like Maxent? You can now use it within SpeciesDistributionToolkit.jl 🎉
We have also added a bunch of dataviz functions, including bivariate and value-suppressing maps. Oh, and conformal prediction for uncertainty quantification. And more coming soon.
poisotlab.github.io/SpeciesDistr...
🌏
We have also added a bunch of dataviz functions, including bivariate and value-suppressing maps. Oh, and conformal prediction for uncertainty quantification. And more coming soon.
poisotlab.github.io/SpeciesDistr...
🌏
March 31, 2026 at 8:02 PM
Do you like Maxent? You can now use it within SpeciesDistributionToolkit.jl 🎉
We have also added a bunch of dataviz functions, including bivariate and value-suppressing maps. Oh, and conformal prediction for uncertainty quantification. And more coming soon.
poisotlab.github.io/SpeciesDistr...
🌏
We have also added a bunch of dataviz functions, including bivariate and value-suppressing maps. Oh, and conformal prediction for uncertainty quantification. And more coming soon.
poisotlab.github.io/SpeciesDistr...
🌏
Stop using MAXENT for invasive niche modeling. Instead embrace mechanisms. Check out this great review on mechanistic niche models for invasives from @erolafenollosa.bsky.social in @ecography.bsky.social 🌐 #ecology #invasivespecies nsojournals.onlinelibrary.wiley.com/doi/full/10....
August 4, 2025 at 3:11 PM
Stop using MAXENT for invasive niche modeling. Instead embrace mechanisms. Check out this great review on mechanistic niche models for invasives from @erolafenollosa.bsky.social in @ecography.bsky.social 🌐 #ecology #invasivespecies nsojournals.onlinelibrary.wiley.com/doi/full/10....
you too are a collection of linear algebra derivable by MaxEnt
June 4, 2026 at 3:27 AM
you too are a collection of linear algebra derivable by MaxEnt
i think "a LLM's forward pass begins as maxent in the Fisher and ends with lower mutual information with the context," which in laymans' terms is "inference in a LLM is a structured process of forgetting what the context was," is an extremely weird way of thinking.
bsky.app/profile/theo...
bsky.app/profile/theo...
i think a lot of us divide the model into "lipschitz work" (rotation and scaling) and "shannon work" (dissipation). mostly the MLP does dissipation. you can _sort_ of think of this model-of-models as having a timelike dimension along the layer index while everything else is spacelike.
October 2, 2026 at 5:21 PM
i think "a LLM's forward pass begins as maxent in the Fisher and ends with lower mutual information with the context," which in laymans' terms is "inference in a LLM is a structured process of forgetting what the context was," is an extremely weird way of thinking.
bsky.app/profile/theo...
bsky.app/profile/theo...
I asked somebody who got hit during the MaxEnt layoffs if at least Mike got shitcanned too and they were like "nope."
I hate it here so much.
I hate it here so much.
July 7, 2026 at 11:09 PM
I asked somebody who got hit during the MaxEnt layoffs if at least Mike got shitcanned too and they were like "nope."
I hate it here so much.
I hate it here so much.
Urban bird diversity conservation plan based on the MaxEnt model and InVEST model: A case study of Jinan, China | Ecological Indicators | www.sciencedirect.co... | #ornithology 🧪🪶 #conservationscience 🌏
April 14, 2025 at 5:01 AM
Urban bird diversity conservation plan based on the MaxEnt model and InVEST model: A case study of Jinan, China | Ecological Indicators | www.sciencedirect.co... | #ornithology 🧪🪶 #conservationscience 🌏
Engaged in "grandpa behavior" (digging up six year out-of-date template code that only exists in a Dropbox folder called "MaxEnt code that works for my friends")
February 11, 2025 at 2:57 PM
Engaged in "grandpa behavior" (digging up six year out-of-date template code that only exists in a Dropbox folder called "MaxEnt code that works for my friends")
by producing only correct answers, we are further away from maxent than the input was, which, unintuitively, is wrong with respct to predicting what a person might write.
September 30, 2026 at 5:22 PM
by producing only correct answers, we are further away from maxent than the input was, which, unintuitively, is wrong with respct to predicting what a person might write.
the MaxEnt of posters
June 30, 2026 at 5:14 PM
the MaxEnt of posters
🌍 DeepMaxent enhances #SpeciesDistributionModelling by integrating neural networks with Maxent, improving predictions even in data-scarce regions. Correcting sampling biases & scaling efficiently, advances biodiversity research & conservation!
👉 More: b-cubed.eu/news/multi-s...
👉 More: b-cubed.eu/news/multi-s...
January 29, 2025 at 2:13 PM
🌍 DeepMaxent enhances #SpeciesDistributionModelling by integrating neural networks with Maxent, improving predictions even in data-scarce regions. Correcting sampling biases & scaling efficiently, advances biodiversity research & conservation!
👉 More: b-cubed.eu/news/multi-s...
👉 More: b-cubed.eu/news/multi-s...
that we can derive attention from MaxEnt and an asymmetric Ising model surely means something
June 22, 2026 at 5:39 AM
that we can derive attention from MaxEnt and an asymmetric Ising model surely means something
What is this, a new essay on entropy? In our year 2024? And it discusses Gibbs and Jaynes and maxent? Embraces a Bayesian interpretation? A Christmas miracle! www.quantamagazine.org/what-is-entr...
What Is Entropy? A Measure of Just How Little We Really Know. | Quanta Magazine
Exactly 200 years ago, a French engineer introduced an idea that would quantify the universe’s inexorable slide into decay. But entropy, as it’s currently understood, is less a fact about the world th...
www.quantamagazine.org
December 22, 2024 at 7:59 AM
What is this, a new essay on entropy? In our year 2024? And it discusses Gibbs and Jaynes and maxent? Embraces a Bayesian interpretation? A Christmas miracle! www.quantamagazine.org/what-is-entr...
Integrating MaxEnt and Random Forest Models to Assess Habitat Suitability of Black-Necked Cranes, A Case Study in Nyingchi City | doi.org/10.1002/ece3... | Ecology and Evolution | #ornithology 🪶
September 9, 2025 at 7:30 AM
Integrating MaxEnt and Random Forest Models to Assess Habitat Suitability of Black-Necked Cranes, A Case Study in Nyingchi City | doi.org/10.1002/ece3... | Ecology and Evolution | #ornithology 🪶
and i think only privacy people think that this doesn't contrast with the "MLPs as key-value store" view, because 'ablating spurious hypotheses' and 'choosing the correct hypotheses' are both steps in the same direction: away from maxent in the Fisher.
October 2, 2026 at 3:54 PM
and i think only privacy people think that this doesn't contrast with the "MLPs as key-value store" view, because 'ablating spurious hypotheses' and 'choosing the correct hypotheses' are both steps in the same direction: away from maxent in the Fisher.
Personally in favor of thermodynamic public opinion (MaxEnt) which is a very good approximation to most polls
June 4, 2025 at 4:47 PM
Personally in favor of thermodynamic public opinion (MaxEnt) which is a very good approximation to most polls
seeing certain fgc content creators saddle up next to that game sucks so much like no, fuck Mike and fuck MaxEnt and their dumb fucking steampunk game
December 4, 2024 at 3:41 PM
seeing certain fgc content creators saddle up next to that game sucks so much like no, fuck Mike and fuck MaxEnt and their dumb fucking steampunk game
Disentangling sampling bias to generate early-warning surfaces of forest ecological sensitivity: An integrated MaxEnt–CatBoost framework www.sciencedirect.com/science/arti...
July 28, 2026 at 10:20 AM
Disentangling sampling bias to generate early-warning surfaces of forest ecological sensitivity: An integrated MaxEnt–CatBoost framework www.sciencedirect.com/science/arti...
"these harms pre-existed X, so X isn't a problem!" is the exact mirror of "these benefits pre-existed X, so X isn't a benefit!"
me and my ass over here thinking in terms of MaxEnt
me and my ass over here thinking in terms of MaxEnt
March 11, 2026 at 7:26 PM
"these harms pre-existed X, so X isn't a problem!" is the exact mirror of "these benefits pre-existed X, so X isn't a benefit!"
me and my ass over here thinking in terms of MaxEnt
me and my ass over here thinking in terms of MaxEnt
New paper out in PRL!
With @artemyte.bsky.social and Sosuke Ito, we combine stochastic thermodynamics with nonequilibrium MaxEnt in large networks to infer entropy dissipation in high-dimensional systems from time series (up to 1000 variables!)
journals.aps.org/prl/abstract...
With @artemyte.bsky.social and Sosuke Ito, we combine stochastic thermodynamics with nonequilibrium MaxEnt in large networks to infer entropy dissipation in high-dimensional systems from time series (up to 1000 variables!)
journals.aps.org/prl/abstract...
February 20, 2026 at 5:07 PM
New paper out in PRL!
With @artemyte.bsky.social and Sosuke Ito, we combine stochastic thermodynamics with nonequilibrium MaxEnt in large networks to infer entropy dissipation in high-dimensional systems from time series (up to 1000 variables!)
journals.aps.org/prl/abstract...
With @artemyte.bsky.social and Sosuke Ito, we combine stochastic thermodynamics with nonequilibrium MaxEnt in large networks to infer entropy dissipation in high-dimensional systems from time series (up to 1000 variables!)
journals.aps.org/prl/abstract...
Back before the GGI acquisition, MaxEnt also just lied publicly about his role on Diesel Legacy, claiming that he was just there to handle netcode.
Credits role and he was Game Director and built the engine, a surprise to no one who knew or worked with him if you even looked at DL for a minute.
Credits role and he was Game Director and built the engine, a surprise to no one who knew or worked with him if you even looked at DL for a minute.
July 7, 2026 at 10:23 PM
Back before the GGI acquisition, MaxEnt also just lied publicly about his role on Diesel Legacy, claiming that he was just there to handle netcode.
Credits role and he was Game Director and built the engine, a surprise to no one who knew or worked with him if you even looked at DL for a minute.
Credits role and he was Game Director and built the engine, a surprise to no one who knew or worked with him if you even looked at DL for a minute.