Tootfinder

Opt-in global Mastodon full text search. Join the index!

No exact results. Similar results found.
@kexpmusicbot@mastodonapp.uk
2026-05-24 12:27:43

🇺🇦 #NowPlaying on KEXP's #VarietyMix
Nilufer Yanya:
🎵 Baby Luv
#NiluferYanya
niluferyanya.bandcamp.com/trac
open.spotify.com/track/6P8b9Tw

@lapizistik@social.tchncs.de
2026-06-24 22:03:00

Sollen sich die gesellschaftlichen Verhältnisse der Wirtschaft oder sich die Wirtschaft den gesellschaftlichen Verhältnissen anpassen?
Letztlich eine politische Frage – die von neoliberalen Kapitalistys dominiert wird.
#pol

@tschfflr@fediscience.org
2026-06-22 15:31:40

Interdisciplinary collaboration can be so eye-opening. I just heard a speaker say off-hand "Oh yeah, this is really new work, from 2017" 😅 I think my colleagues in #NLProc wouldn't even touch such historic sources #academicChatter

@radioeinsmusicbot@mastodonapp.uk
2026-06-23 15:17:35

🇺🇦 Auf radioeins läuft...
Neal Francis:
🎵 Hide
#NowPlaying #NealFrancis
open.spotify.com/track/197QlBc

@villavelius@mastodon.online
2026-05-23 09:24:51

Neoliberal capitalist ideology militates against fairness, and consequently undermines the support base of democracy.

@arXiv_csCR_bot@mastoxiv.page
2026-07-24 07:58:05

Leaky Language Models: Stealing Architecture and Inference Optimizations via Per-Token Timing
Sadegh Majidi, Niloofar Mireshghallah, Kazem Taram
arxiv.org/abs/2607.20723 arxiv.org/pdf/2607.20723 arxiv.org/html/2607.20723
arXiv:2607.20723v1 Announce Type: new
Abstract: This work presents LeakyLMs, a set of attacks that leak proprietary model, architecture, and deployment information from production language models. LeakyLMs is the first to demonstrate that key model and deployment details can be inferred using only token generation timing, even when interacting through remote APIs. LeakyLMs introduces two core attacks. The first attack targets inference optimizations and deployment strategies. For example, our attack detects whether a provider uses speculative decoding, a widely deployed inference-time optimization, and further identifies the context length of the draft model used in the pipeline. Our measurements show that Google Gemini Flash 2.5 uses speculative decoding with a draft context window of approximately 128K tokens. The second attack recovers key architectural properties, including the number of transformer layers, hidden dimension size, and number of attention heads. To achieve this, LeakyLMs builds a detailed and accurate model of token-generation timing on modern NVIDIA GPUs, characterizing how latency scales with model configuration and hardware parameters. The attack then performs a search over the architecture space using this timing model. In experiments with Llama models, the near-correct architectural configuration appears in the top-10 guesses more than 90% of the time.
toXiv_bot_toot

@PaulWermer@sfba.social
2026-07-21 20:00:47

Ah, yes - today's oligopoly learning from past abuses that were never addressed because of the #NeoliberalIdeology. (Amazon is not the only company to use it's volume to drive up costs or restrict product availability for competitors - it's been happening for years by large retail brands.)
‘We were at their mercy’: inside the Amazon tactics that hiked prices across the …

@paulwermer@sfba.social
2026-07-21 20:00:47

Ah, yes - today's oligopoly learning from past abuses that were never addressed because of the #NeoliberalIdeology. (Amazon is not the only company to use it's volume to drive up costs or restrict product availability for competitors - it's been happening for years by large retail brands.)
‘We were at their mercy’: inside the Amazon tactics that hiked prices across the …

@BBC6MusicBot@mastodonapp.uk
2026-06-25 13:51:38

🇺🇦 #NowPlaying on #BBC6Music's #CraigCharles
Nilüfer Yanya:
🎵 Method Actor
#NilüferYanya
niluferyanya.bandcamp.com/trac

@radioeinsmusicbot@mastodonapp.uk
2026-05-22 08:48:25

🇺🇦 Auf radioeins läuft...
Neal Francis:
🎵 Don't Want You To Know
#NowPlaying #NealFrancis
open.spotify.com/track/6FGDNEM