Tootfinder

The Allen Institute for AI open sources OLMo, or "Open Language MOdels", and its data set Dolma; OLMo was created with Harvard, AMD, Databricks, and others (Kyle Wiggers/TechCrunch)
https://techcrunch.com/2024/02/01/ai2-

AI2 open sources text-generating AI models -- and the data used to train them | TechCrunch
AI2, the AI research institute, has released several text-generating models in open source, along with the data used to train them.

Models All The Way Down #tceajmg #edtech

New blog post (not April fools!):
“Hidden models and latent compression in community detection”
This an overdue post on a joint work with Alec Kirkley published last year.
https://skewed.de/tiago/posts/hidden-models/

https://www.middleeasteye.net/news/exclusive-israel-planning-ring-checkpoints-prevent-men-fleeing-rafah

Exclusive: Israel planning ring of checkpoints to prevent men from fleeing Rafah
Israel is setting up 'complex network' of checkpoints to prevent 'military age' men from fleeing Rafah assault, senior western official tells MEE

World's
Tiniest
Violin
OpenAI accuses New York Times of hacking AI models in copyright lawsuit
https://cointelegraph.com/news/openai-new-york-times-hacking-ai-models

Check out this story about the methods of #AI model training. The stacking. Etc. A qualitative critique.
"It’s only by looking at datasets that we can get a better sense of how AI models work, and the gaps, errors, and biases that can emerge."

Navigating Brain Language Representations: A Comparative Analysis of Neural Language Models and Psychologically Plausible Models
Yunhao Zhang, Shaonan Wang, Xinyi Dong, Jiajun Yu, Chengqing Zong
https://arxiv.org/abs/2404.19364 https://arxiv.org/pdf/2404.19364
arXiv:2404.19364v1 Announce Type: new
Abstract: Neural language models, particularly large-scale ones, have been consistently proven to be most effective in predicting brain neural activity across a range of studies. However, previous research overlooked the comparison of these models with psychologically plausible ones. Moreover, evaluations were reliant on limited, single-modality, and English cognitive datasets. To address these questions, we conducted an analysis comparing encoding performance of various neural language models and psychologically plausible models. Our study utilized extensive multi-modal cognitive datasets, examining bilingual word and discourse levels. Surprisingly, our findings revealed that psychologically plausible models outperformed neural language models across diverse contexts, encompassing different modalities such as fMRI and eye-tracking, and spanning languages from English to Chinese. Among psychologically plausible models, the one incorporating embodied information emerged as particularly exceptional. This model demonstrated superior performance at both word and discourse levels, exhibiting robust prediction of brain activation across numerous regions in both English and Chinese.

Rechtsaußen-#Demo will in #Regensburg marschieren und klagt
Mittelstands-Demo mit „Hauptredner“ #Aiwanger distanziert sich

Very important:
"Investigating #trainingsets is an essential avenue to understanding how #generativeAI models work; the ways they see and re-create the world."

JFrog says it found around a hundred malicious ML models on Hugging Face, some of which can backdoor users' machines (Bill Toulas/BleepingComputer)
https://www.bleepingcomputer.com/news/security/malicious-ai-models-on…