Manacá-1B-Instruct-2: Reproducible Safety Alignment for a Native Brazilian-Portuguese LLM. Negative DPO Results, a Controlled Composition Scan, and a Silent Failure Mode of the LLM Judge
Paper recorded by Signals 4 on 2026-10-14 in cs.CL. Abstract reproduced from arXiv; link to the original below.
Published 2026-10-14 on arXiv · recorded by Signals 4 on 2026-10-15
Category: cs.CL · 自然语言处理 · first seen 2026-10-15
Abstract
Brazilian Portuguese remains underserved by open-source language models, and the few that are safety-aligned rarely document what did not work or validate comparisons with statistical significance. We present Manacá-1B-Instruct-2, the second-generation post-training alignment of Manacá-1B (1.72B parameters, trained from scratch for PT-BR), with an SFT corpus of 291,223 examples. We document three
Read on arXiv →
Cite this page: Manacá-1B-Instruct-2: Reproducible Safety Alignment for a Native Brazilian-Portuguese LLM. Negative DPO Results, a Controlled Composition Scan, and a Silent Failure Mode of the LLM Judge: the #242 most recent of 660 cs.CL papers we have recorded (as of 2026-10-14). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/manac-1b-instruct-2-reproducible-safety-alignment-for-a-native-brazilian-portugu.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable:
papers.json
Get 4 AI signals a day by email — free.
Get 4 AI signals a day by email — free