#interpretable
LightNews LightNews
joewintergreen.com
like the amount of times i have somebody telling me i said something horrible, and i'm like "where", and they screenshot something i said that isn't remotely interpretable as what they think it was... it does explain a lot if they genuinely can't read and only think they can
sarahjeong.bsky.social sarah jeong @sarahjeong.bsky.social · 2d
ever since I learned about three-cueing I've developed infinitely more patience for replies on social media. mfers literally do not know how to read. people are walking around conjuring random meanings into words they don't know, and they don't know a lot of words. it's crazy
November 12, 2025 at 6:48 AM
jcamthrash.bsky.social
Decoding emergent properties of microbial community functions through subcommunity observations and interpretable machine learning academic.oup.com/ismej/articl... #jcampubs
Decoding emergent properties of microbial community functions through subcommunity observations and interpretable machine learning
Abstract. The functions of microbial communities, including substrate conversion and pathogen suppression, arise not as a simple sum of individual species’
academic.oup.com
November 11, 2025 at 3:07 PM
pseudomanifold.topology.rocks
Big graph, little compute?

In our #NeurIPS2025 paper we propose geometry-aware edge-contraction-based pooling methods for GNNs.

Our methods preserve graph structure, make interpretable pooling choices, and ensure robust performance at downstream tasks.

🧵1/n
An overview of our edge pooling method.
November 11, 2025 at 3:48 PM
entzuendung.bsky.social
You can add more words to try to be more precise, but you just end up widening the interpretable surface. You end up making more space for misunderstanding. Even "misunderstanding" is interesting. As if there was a real underlying, definitive, unambiguous understanding to be attained.
November 9, 2025 at 6:39 PM
pettertornberg.com
We introduce a Computational Turing Test — a validation framework that compares human and LLM text using:

🕵️‍♂️ Detectability — can an ML classifier tell AI from human?

🧠 Semantic fidelity — does it mean the same thing?

✍️ Interpretable linguistic features — style, tone, topics.
November 7, 2025 at 11:13 AM
wiredferret.bsky.social
He demanded his phone and now he's looking at TV static and describing what he sees. Fish, matrix, escher...

Kaleidescope is overstimulating.
Mandelbrot zoom not interpretable enough, has its own meanings too firmly.
November 13, 2025 at 1:29 PM
ulrikeluxburg.bsky.social
Here is a formal impossibility result for XAI: Informative Post-Hoc Explanations Only Exist for Simple Functions. I'll give an online presentation about this work next tuesday in @timvanerven.nl 's Theory of Interpretable AI Seminar:

arxiv.org/abs/2508.11441

tverven.github.io/tiai-seminar/
November 7, 2025 at 6:25 AM
rbregier.bsky.social
Meet Anny, our Free (Apache 2.0) and Interpretable Human Body Model for all ages.
Anny is built upon #MakeHuman and enables achieving SOTA performance in Human Mesh Recovery.
ArXiv: arxiv.org/abs/2511.03589
Demo: anny-demo.europe.naverlabs.com
Code: github.com/naver/anny
November 6, 2025 at 10:59 AM
myeyesviolet.bsky.social
this is very confusing honestly, she's made no secret about the meaning of the song... there's nothing about a fake "jewel" and having to make your own happiness that feels interpretable to me as her happiness coming from the man who put her on blast on a podcast EVEN IF IT WAS "METAL AS HELL"
scoopofsherbet.bsky.social Shelby @scoopofsherbet.bsky.social · 7d
Dude I hope swifties get everything thats coming they have taken the Kayla Nicole shit too far, someone needs to tell Kayla its not a good thing the sky is opalite or something
November 7, 2025 at 12:55 AM
mlolab.bsky.social
Working with multi-omic or spatial data?

Meet MOFA-FLEX — a modular framework for interpretable factor analysis. Build custom models, select sparsity priors, leverage domain knowledge, integrate spatial information and explore results!

Here’s a quick tutorial 👇
November 7, 2025 at 10:24 AM
ricoramirez.bsky.social
Great to see our patient map of heart failure across patient cohorts, scales, and techs out.

Hope that our work shows how we can transform molecular data into interpretable tissue features that describe remodeling and associate with clinical outcomes.

www.nature.com/articles/s41...
A cross-study transcriptional patient map of heart failure defines conserved multicellular coordination in cardiac remodeling - Nature Communications
Cardiac tissue remodeling in heart failure is driven by interactions between multiple cell types, but existing studies have not fully captured these coordinated responses. Here, the authors show that ...
www.nature.com
November 6, 2025 at 1:18 PM
jarriel-desgracia.bsky.social
Mil Taras es el nuevo EP de Jarriel y Sz. Hoy os presentamos el primer tema titulado: Calor. Link en Bio! 🔗

#musica #rap #beat #sevilla #hiphop #interpretable #lofi #nostalgia
November 6, 2025 at 5:48 PM
envdatascience.bsky.social
New article!

From winter storm thermodynamics to wind gust extremes: discovering interpretable equations from data

👉 https://bit.ly/4oPzZSN
✍️ Frederick Iat-Hin Tam, Fabien Augsburger, Tom Beucler (@fgse-unil.bsky.social, @unil.bsky.social )

@climformatics.bsky.social #CI2025 #thermodynamics #wind
November 5, 2025 at 11:02 AM
croquebitume.bsky.social
Vrai que l'affirmation de départ est trop largement interprétable ☝️😅🤦😂🤣
November 4, 2025 at 3:21 PM
asaakyan.bsky.social
Writing quality reward model scores are associated with both creativity and pragmaticality judgements, but are not interpretable. LLM-judge can replicate some expert novelty judgements but struggle with identifying non-pragmatic expressions.
November 4, 2025 at 3:08 PM
aida-alvera.bsky.social
📣 🌊 Call for abstracts #EGU26 !!! Please consider our session:

ITS1.9/OS4.1
Machine Learning for Ocean Science
meetingorganizer.copernicus.org/EGU26/abstra...

Deadline 15 January 2026 (1 December for travel grant applications) @brajard.bsky.social @rachelfurner.bsky.social @redouanelg.bsky.social
A slide showing a blue surface with waves and superposed 0s and 1s on the left, and text describing the session and details for submission: ITS1.9/OS4.1 Machine Learning for Ocean Science https://meetingorganizer.copernicus.org/EGU26/sessionprogramme/5869# Session abstract: Machine learning (ML) methods have emerged as powerful tools to tackle various challenges in ocean science, encompassing physical oceanography, biogeochemistry, and sea ice research. This session aims to explore the application of ML methods in ocean science, with a focus on advancing our understanding and addressing key challenges in the field. Our objective is to foster discussions, share recent advancements, and explore future directions in the field of ML methods for ocean science. A wide range of machine learning techniques can be considered including supervised learning, unsupervised learning, interpretable techniques, and physics-informed and generative models. The applications to be addressed span both observational and modeling approaches. Observational approaches include for example: - Identifying patterns and features in oceanic fields - Filling observational gaps of in-situ or satellite observations - Inferring unobserved variables or unobserved scales - Automating quality control of data - Modeling approaches can address (but are not restricted to): - Designing new parameterization schemes in ocean models - Emulating partially or completely ocean models - Parameter tuning and model uncertainty The session also welcomes submissions at the interface between modeling and observations, such as data assimilation, data-model fusion, or bias correction. Researchers and practitioners working in the domain of ocean science, as well as those interested in the application of ML methods, are encouraged to attend and participate in this session.
November 5, 2025 at 9:03 AM
hockeyviz.com
Point systems are abstract; you can't get a feel for what they mean without experience. Win-loss ratings have a clearly interpretable meaning by themselves even for new fans, even before you add details like "how are the other teams doing" or "where are we in the season".
November 3, 2025 at 10:11 PM
jrdg42.bsky.social
Es parcialmente interpretable como rendirse, pero también un acto de resistencia. No es fácil moralmente. Pero al menos espero que mi presencia y actos, por fútiles que sean, valen la pena. Es una pizca de mi activismo.
November 2, 2025 at 1:19 PM
stephenjudkins.bsky.social
I like how this pitch is transparently interpretable as "we will squeeze out every last dollar from America's most degenerate scumbags, invest in us"
November 1, 2025 at 3:09 AM
davidcrespo.bsky.social
this is pretty good and doesn't push the red herring of black box interpretability — transparency and control are social problems primarily
xcancel.com/davidcrespo/...
conputer dipshit @davidcrespo 19 Nov 2018 the one point I’ll add that I don’t see from the thread is that the vast majority of what is hyped as “machine learning” is just regressions and decision trees, which means the interpretability/black box problem is overblown bc those are interpretable conputer dipshit @davidcrespo 19 Nov 2018 this suggests that what is at issue is not so much interpretability in an absolute sense but rather whether the systems making decisions about are interpretable *by us* and accountable *to us* Nov 19, 2018 · 2:57 AM UTC conputer dipshit @davidcrespo 19 Nov 2018 the regression model used to calculate my social credit score may well be perfectly interpretable by the data scientists who built it, but that doesn’t mean I have any way of finding out or changing it
November 2, 2025 at 2:44 PM
amydwillis.bsky.social
Happy Friday from the StatDivLab!

We just made updates to radEmu to make it faster & more stable than ever.

Thanks to all our users -- we're excited to keep supporting it.

github.com/statdivlab/r...

Interpretable and robust microbial differential abundance, woohoo!
GitHub - statdivlab/radEmu
Contribute to statdivlab/radEmu development by creating an account on GitHub.
github.com
October 31, 2025 at 8:48 PM
natmethods.nature.com
CytoTRACE 2: an interpretable deep-learning method that uses scRNA-seq data to predict absolute developmental potential across datasets.

www.nature.com/articles/s41...
Improved reconstruction of single-cell developmental potential with CytoTRACE 2 - Nature Methods
CytoTRACE 2 is an interpretable deep learning framework that leverages single-cell RNA sequencing data to predict absolute developmental potential across datasets.
www.nature.com
October 29, 2025 at 3:14 PM
dsquintana.bsky.social
Who’s gonna pay @rpsychologist.com to make a mega tool converting all g*power analysis types to an interpretable web tool like this?
rpsychologist.com Kristoffer Magnusson @rpsychologist.com · 16d
New release of PowerLMM.js! Browser-based power analysis for longitudinal models with dropout.

Now includes:
- Power analysis summary report
- Reproducible & shareable configs (URL/JSON)
- Calculations validated against R
- Hypothesis region visualization

powerlmmjs.rpsychologist.com
October 28, 2025 at 3:18 PM
hbcproject.bsky.social
📢 Paper update!

A revised version of “Predicting outcomes of smoking cessation interventions in novel scenarios using ontology-informed, interpretable machine learning” is now published.

Read it here 👉 wellcomeopenresearch.org/articles/8-5...

Many thanks to everyone who contributed!
wellcomeopenresearch.org
October 29, 2025 at 6:21 PM
genomicspc.bsky.social
🌍 We are excited to launch the European genomics #preprint club!

Two preprints to discuss tomorrow:

1️⃣ EnhancerDetector: Enhancer Discovery from Human to Fly via Interpretable Deep Learning

2️⃣ Inferring Gene-Regulatory Networks Using Epigenomic Priors

Contact us if you are interested in joining.
October 29, 2025 at 7:18 PM