Tootfinder

I can't get over how bad the docker ecosystem is.
"we provide a docker image, just use that!"
*install container image & run it - it doesn't work*
"oops, we didn't test it"

🧠 The reasoning: text tokens are discrete, drawn from a fixed vocabulary of ~50,000 entries, each mapping to a fixed point in embedding space.
Image tokens are continuous vectors of a thousand floating point numbers, so they can pack far more information into the same space.
🎮 The test: a tower-defense game where the AI plays in JSON mode without graphics, run for hours with two context variants — the same game state as a plain text string, and as an image rendered via PixelPipe.…

I set the alarm to 6am and we started at 7am. My wife cycled a small morning loop (like 45min) and I did a bit larger loop to test the 30 tooth Sprocket on longer climbs.
*Learnings today:*
- mosquitos can be a real motivator to push harder
- the sprocket feels good
- the #garmin incident detection works
- 1st aid kit is a good investment
- good medical coverage in …

🖼️ Can images replace text as LLM context? A #DeepSeek paper claims one image token carries ~10 text tokens of information at close to 100% accuracy, with 59-70% cost reduction reported by PixelPipe. ThePrimeagen put it to the test. #AI

i vastly prefer all the erotica I accidentally see on Fedi pass the Bechdel test. But also be CW'd if there's an image, because not everyone on the train behind me consented to being jumpscared by boobs.
#FediMeta #JustGirlyThings

What causes the image preview to fail when sharing a link on Mastodon? Is there a way to test prior to posting?

Replaced article(s) found for eess.AS. https://arxiv.org/list/eess.AS/new
[1/1]:
- Unifying Diarization, Separation, and ASR with Multi-Speaker Encoder
Muhammad Shakeel, Yui Sudo, Yifan Peng, Chyi-Jiunn Lin, Shinji Watanabe
https://arxiv.org/abs/2508.20474 https://mastoxiv.page/@arXiv_eessAS_bot/115110974009150613
- CALM: Joint Contextual Acoustic-Linguistic Modeling for Personalization of Multi-Speaker ASR
Muhammad Shakeel, Yosuke Fukumoto, Chikara Maeda, Chyi-Jiunn Lin, Shinji Watanabe
https://arxiv.org/abs/2601.22792 https://mastoxiv.page/@arXiv_eessAS_bot/116000207024295325
- How Much Does Machine Identity Matter in Anomalous Sound Detection at Test Time?
Kevin Wilkinghoff, Keisuke Imoto, Zheng-Hua Tan
https://arxiv.org/abs/2602.16253 https://mastoxiv.page/@arXiv_eessAS_bot/116096185732811365
- LMU-Based Sequential Learning and Posterior Ensemble Fusion for Cross-Domain Infant Cry Classific...
Niloofar Jazaeri, Hilmi R. Dajani, Marco Janeczek, Martin Bouchard
https://arxiv.org/abs/2603.02245 https://mastoxiv.page/@arXiv_eessAS_bot/116169771215037748
- Adapting a Text-to-Audio Model for Room Impulse Response Generation
Kirak Kim, Sungyoung Kim
https://arxiv.org/abs/2603.09708 https://mastoxiv.page/@arXiv_eessAS_bot/116209762413602825
- Repurposing Image Diffusion Models for Training-Free Music Style Transfer on Mel-spectrograms
Heehwan Wang, Joonwoo Kwon, Sooyoung Kim, Jungwoo Seo, Shinjae Yoo, Yuewei Lin, Jiook Cha
https://arxiv.org/abs/2411.15913 https://mastoxiv.page/@arXiv_csSD_bot/113548024475383386
- DeePen: Penetration Testing for Audio Deepfake Detection
M\"uller, Kawa, Stan, Doan, Jung, Choong, Sperl, B\"ottinger
https://arxiv.org/abs/2502.20427 https://mastoxiv.page/@arXiv_csCR_bot/114097333876265997
- Re-evaluating Minimum Bayes Risk Decoding for Automatic Speech Recognition
Yuu Jinnai
https://arxiv.org/abs/2510.19471 https://mastoxiv.page/@arXiv_csCL_bot/115422969877240889
- Aliasing-Free Neural Audio Synthesis
Yicheng Gu, Junan Zhang, Chaoren Wang, Jerry Li, Zhizheng Wu, Lauri Juvela
https://arxiv.org/abs/2512.20211 https://mastoxiv.page/@arXiv_csSD_bot/115773521971327576
- TiCo: Time-Controllable Spoken Dialogue Model
Kai-Wei Chang, Wei-Chih Chen, En-Pei Hu, Hung-yi Lee, James Glass
https://arxiv.org/abs/2603.22267 https://mastoxiv.page/@arXiv_csCL_bot/116283643505371784
toXiv_bot_toot

In other news, my first 15x11" (roughly DIN A3) print came out with only minor issues, none of them show stoppers and I know what needs to be done to improve. So this is just a test print and it's a bit hard to capture, but the overall presence & level of detail is so amazing at this size! Hard to go back to smaller sizes now :)
To get here, I've been doing countless tests of different recipes and hardware (e.g. I had to disassemble my UV light to create a more uniform/unfocused li…

Our #VFXPipeline, database and renderfarm also supports shot numbers with emojis. Need to find my post about this from a few years ago 😁
Maybe I should test it again to make sure it still works. Back then, all I had to do was disable a regular expression that enforced alphanumeric characters.