Closing the series with what is not a research line but makes them possible: open models served as Slurm jobs for others to use, hardware validation, the tests prior to enabling containers …
Detecting grooming, harassment or threats in a minor's conversations without the conversation leaving the phone. The line generates half a million synthetic examples, fine-tunes a …
Almost 300 GGUF files from dozens of families, served with llama.cpp and scored on function calling, code and long context, with speed and energy per card. The line that reserves the most …
Several small open models deliberate to build prompts that make gpt-oss-120b give in. The line measures how much deliberating adds over attacking alone, and devotes as much effort to the …
A 3-billion-parameter Llama 3.2, LoRA fine-tuning and one question: what dose of safety data is needed for it to refuse what is harmful without refusing everything. The line tries fixed …
Four months of the compute cluster of the ANTS group at the University of Murcia, reconstructed from the Slurm jobs themselves and anonymised: what is really computed and what engineering is …
AI-generated translation
Everything on this site so far was written in Spanish and translated into English by an AI. The translation has not been reviewed by a person: the Spanish version is the one that counts.
