Tootfinder

BPO giant Conduent confirms data breach impacts 10.5 million people
https://www.bleepingcomputer.com/news/security/bpo-giant-conduent-confirms-data-breach-impacts-105-million-people/

BPO giant Conduent confirms data breach impacts 10.5 million people
American business services giant Conduent has confirmed that a 2024 data breach has impacted over 10.5 million people, according to notifications filed with the US Attorney General's offices.

from my link log —
Conversions in Ruby are a mess.
https://github.com/sampersand/blog/blob/master/conversions.md
saved 2025-10-30 https:…

blog/conversions.md at master · sampersand/blog
Place to put articles before I finally do make a blog - sampersand/blog

Convergence and Divergence of Language Models under Different Random Seeds
Finlay Fehlauer (ETH Zurich), Kyle Mahowald (University of Texas at Austin), Tiago Pimentel (ETH Zurich)
https://arxiv.org/abs/2509.26643

Convergence and Divergence of Language Models under Different Random Seeds
In this paper, we investigate the convergence of language models (LMs) trained under different random seeds, measuring convergence as the expected per-token Kullback--Leibler (KL) divergence across seeds. By comparing LM convergence as a function of model size and training checkpoint, we identify a four-phase convergence pattern: (i) an initial uniform phase, (ii) a sharp-convergence phase, (iii) a sharp-divergence phase, and (iv) a slow-reconvergence phase. Further, we observe that larger mode…

QGIS :qgis: to (Geo)Pandas follow-up:
The conversation around Looking for better ways to convert between QGIS VectorLayer and (Geo)DataFrame is continuing over at https://fosstodon.org/@underdarkGIS/115442614331293320

QGIS to (Geo)Pandas follow-up
The conversation around Looking for better ways to convert between QGIS VectorLayer and (Geo)DataFrame is continuing over at What I’ve learned so far: QgsVectorLayer.as_geopandas() has landed…

ENQUETE. Corruption, favoritisme, décisions illégales... La Cour des comptes dénonce les dérives des chambres d'agriculture
https://www.franceinfo.fr/enquetes-francei

ENQUETE. Corruption, favoritisme, décisions illégales... La Cour des comptes dénonce les dérives des chambres
Dans un rapport que la cellule investigation de Radio France a pu consulter en exclusivité, la Cour des comptes pointe "irrégularités" et "dysfonctionnements" au sein des chambres d'agriculture françaises. L'État, maintes fois alerté, est accusé d'avoir trop souvent laissé faire.

CEVAW Conversations
Great Australian Pods Podcast Directory: #GreatAusPods

'Confirmation of Kremlin state policy' — Ukraine slams Russia's withdrawal from torture prevention convention: https://benborges.xyz/2025/09/29/confirmation-of-kremlin-state-policy.html

EZ-Sort: Efficient Pairwise Comparison via Zero-Shot CLIP-Based Pre-Ordering and Human-in-the-Loop Sorting
Yujin Park, Haejun Chung, Ikbeom Jang
https://arxiv.org/abs/2508.21550

EZ-Sort: Efficient Pairwise Comparison via Zero-Shot CLIP-Based Pre-Ordering and Human-in-the-Loop Sorting
Pairwise comparison is often favored over absolute rating or ordinal classification in subjective or difficult annotation tasks due to its improved reliability. However, exhaustive comparisons require a massive number of annotations (O(n^2)). Recent work has greatly reduced the annotation burden (O(n log n)) by actively sampling pairwise comparisons using a sorting algorithm. We further improve annotation efficiency by (1) roughly pre-ordering items using the Contrastive Language-Image Pre-trai…

BALF: Budgeted Activation-Aware Low-Rank Factorization for Fine-Tuning-Free Model Compression
David Gonz\'alez Mart\'inez
https://arxiv.org/abs/2509.25136 https://

BALF: Budgeted Activation-Aware Low-Rank Factorization for Fine-Tuning-Free Model Compression
Neural network compression techniques typically require expensive fine-tuning or search procedures, rendering them impractical on commodity hardware. Inspired by recent LLM compression research, we present a general activation-aware factorization framework that can be applied to a broad range of layers. Moreover, we introduce a scalable budgeted rank allocator that allows flexible control over compression targets (e.g., retaining 50% of parameters) with no overhead. Together, these components f…