Lobste.rs
Ir al canal en Telegram
1 792
Suscriptores
-224 horas
Sin datos7 días
+2930 días
Carga de datos en curso...
Canales Similares
Nube de Etiquetas
Menciones Entrantes y Salientes
---
---
---
---
---
---
Atraer Suscriptores
agosto '26
agosto '26
+52
en 3 canales
julio '26
+57
en 7 canales
Get PRO
junio '26
+68
en 8 canales
Get PRO
mayo '26
+78
en 8 canales
Get PRO
abril '26
+81
en 6 canales
Get PRO
marzo '26
+56
en 10 canales
Get PRO
febrero '26
+69
en 7 canales
Get PRO
enero '26
+53
en 4 canales
Get PRO
diciembre '25
+42
en 7 canales
Get PRO
noviembre '25
+50
en 6 canales
Get PRO
octubre '25
+52
en 7 canales
Get PRO
septiembre '25
+45
en 4 canales
Get PRO
agosto '25
+61
en 8 canales
Get PRO
julio '25
+52
en 7 canales
Get PRO
junio '25
+58
en 5 canales
Get PRO
mayo '25
+62
en 7 canales
Get PRO
abril '25
+66
en 7 canales
Get PRO
marzo '25
+65
en 6 canales
Get PRO
febrero '25
+55
en 5 canales
Get PRO
enero '25
+58
en 6 canales
Get PRO
diciembre '24
+100
en 10 canales
Get PRO
noviembre '24
+85
en 6 canales
Get PRO
octubre '24
+104
en 4 canales
Get PRO
septiembre '24
+86
en 4 canales
Get PRO
agosto '24
+82
en 3 canales
Get PRO
julio '24
+69
en 6 canales
Get PRO
junio '24
+67
en 5 canales
Get PRO
mayo '24
+55
en 3 canales
Get PRO
abril '24
+49
en 3 canales
Get PRO
marzo '24
+52
en 7 canales
Get PRO
febrero '24
+60
en 3 canales
Get PRO
enero '24
+77
en 5 canales
Get PRO
diciembre '23
+64
en 4 canales
Get PRO
noviembre '23
+20
en 3 canales
Get PRO
octubre '23
+21
en 4 canales
Get PRO
septiembre '23
+412
en 0 canales
| Fecha | Crecimiento de Suscriptores | Menciones | Canales | |
| 29 agosto | +1 | |||
| 28 agosto | +1 | |||
| 27 agosto | +1 | |||
| 26 agosto | +2 | |||
| 25 agosto | +4 | |||
| 24 agosto | 0 | |||
| 23 agosto | 0 | |||
| 22 agosto | +1 | |||
| 21 agosto | +2 | |||
| 20 agosto | +1 | |||
| 19 agosto | +4 | |||
| 18 agosto | +3 | |||
| 17 agosto | +1 | |||
| 16 agosto | +1 | |||
| 15 agosto | 0 | |||
| 14 agosto | +1 | |||
| 13 agosto | +6 | |||
| 12 agosto | +4 | |||
| 11 agosto | +3 | |||
| 10 agosto | +3 | |||
| 09 agosto | +2 | |||
| 08 agosto | +1 | |||
| 07 agosto | +1 | |||
| 06 agosto | +1 | |||
| 05 agosto | +1 | |||
| 04 agosto | +3 | |||
| 03 agosto | +2 | |||
| 02 agosto | 0 | |||
| 01 agosto | +2 |
Publicaciones del Canal
| 2 | Bug blindness
Comments
via danluu.com via nolan | 54 |
| 3 | Omnipresent availability risks in cloud software
Comments
via surfingcomplexity.blog via azhenley | 93 |
| 4 | Functional State Machines in Rust: Typestate and Newtype Patterns
Comments
via dl.acm.org via anex9d | 103 |
| 5 | What GLM-5.3 Flash running on Chinese hardware actually means
Comments
via martinalderson.com by martinald | 120 |
| 6 | Introducing ReactOS 0.4.16
Comments
via reactos.org via bb010g | 110 |
| 7 | Canonical-basis realignment for Transformer LLMs: every hidden axis becomes independently measurable and controllable
The code essentially gives you a way to rotate a Transformer's internal coordinate system into a canonical basis that aligns with its own weight matrices in a lossless way. By absorbing the normalization gains directly into the adjacent weights and using orthogonal matrices built from the singular vectors of the model, you can transform architectures like Qwen or Pythia without altering their outputs or perplexity scores.Applying this transform reveals the actual hidden geometric structures operating inside the network. Once the model is rotated into this new perspective, you can see its internal mechanisms that were previously opaque. The authors found things like a bipolar oscillator where specific axes form inhibitory pairs that fire against each other in perfect opposition. They also observed a kind of rhythmic respiration across layers where the model alternates between absorbing knowledge and filtering it. On top of that, it exposed a homeostatic defense mechanism that aggressively erases any localized perturbations within just a couple of layers.Practically speaking, researchers now have a powerful lens for mapping out how models actually do reasoning. For example, it turns out that the effective rank of the correlation matrix in a half billion parameter model might be as low as eleven independent patterns. Reframing how we look at the internal activations of language models provides a standardized way to study their underlying architecture.Comments
via github.com via Yogthos | 110 |
| 8 | Wrapping GTK4 in 800 lines of Clojure with Jolt
Comments
via yogthos.net by Yogthos | 130 |
| 9 | Does Computer Science Need Computers?
Comments
via quantamagazine.org via m-hodges | 167 |
| 10 | Creepy crawlies
Comments
via people.kernel.org via gmem | 155 |
| 11 | Civic Hygiene (2013)
Comments
via shkspr.mobi via lim | 162 |
| 12 | Debian and the sirens
Comments
via joeyh.name via danlamanna | 153 |
| 13 | AI Revives a Dead Sleep Company
Comments
via josephspurrier.com via carlana | 150 |
| 14 | On 'scaling up' and being the right size
Comments
via tzovar.as via koala | 163 |
| 15 | Reviving An SD Card With Shorted Capacitors
Comments
via hackaday.com via metahost | 176 |
| 16 | Curvature Beziers
Comments
via acko.net via fanf | 163 |
| 17 | InferenceFS: Never worry about data again! (Again!)
Comments
via github.com via linkdd | 158 |
| 18 | Debian Votes To Allow "Responsible Use Of Generative AI"
Comments
via phoronix.com via laktak | 158 |
| 19 | Parsing the Infamous Japanese Postal CSV
Comments
via dampfkraft.com via lim | 165 |
| 20 | Being kicked out of the tech industry
Comments
via jacky.wtf by jalcine | 173 |
