arXiv Computation and Language October 15, 2026 By Bruno Leonardo Santos Menezes, Carlos Leonardo Souza Cardoso, Fabio Andr\'e Machado Porto

Manac\'a-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

Read the original on arXiv Computation and Language →

The Flow has not summarised this story yet — read it at arXiv Computation and Language.

arXiv AI

Aug 18

Toward Better Assessment of LLMs' Performance in Clinical Error Detection

arXiv:2608. 16643v1 Announce Type: cross Abstract: Automated detection of errors in clinical documentation is a promising application of large language models (LLMs), yet decisions to deploy such models rest on benchmarks that evaluate each clinical note in isolation.

By Yifan Zhang, Rahmatollah Beheshti

llms nlp benchmarks safety

arXiv Machine Learning

Aug 4

Efficiency vs. Alignment: Investigating Safety and Fairness Risks in Parameter-Efficient Fine-Tuning of LLMs

arXiv:2511. 00382v2 Announce Type: replace-cross Abstract: Organizations increasingly adapt Large Language Models (LLMs) from public repositories such as HuggingFace to downstream tasks.

By Mina Taraghi, Yann Pequignot, Amin Nikanjam, Mohamed Amine Merzouk, Foutse Khomh

llms fine-tuning safety

arXiv Machine Learning

Sep 30

Storage Is Not Strategy: State-Conditioned Support Control for LLM Unlearning

arXiv:2609.37858v1 Announce Type: new Abstract: Many localized large language model (LLM) unlearning methods select a small parameter subset from a localization signal and keep it fixed during optimi...

By Tianhao Qian, Ziming Hong, Chongyang Gao, Kezhen Chen, Lixu Wang

llms fine-tuning benchmarks

arXiv Machine Learning

Jul 30

Constitutional Midtraining: Content Presence Drives Alignment Gains

arXiv:2607. 26654v1 Announce Type: cross Abstract: Post-training alignment is often shallow, eroding under fine-tuning.

By Desiree Cho, Cameron Tice, Bernie Hogan, Hunar Batra, Puria Radmard, Jun Zhao, Nigel Shadbolt

fine-tuning benchmarks safety

arXiv AI

Jul 3

Breaking Safety at the Token Boundary: How BPE Tokenization Creates Exploitable Gaps in LLM Alignment

arXiv:2607. 01239v1 Announce Type: cross Abstract: Character-level perturbations bypass safety alignment in modern LLMs despite leaving prompts human-readable.

By Tung-Ling Li, Hongliang Liu, Yuhao Wu

llms nlp reinforcement-learning fine-tuning multimodal safety

arXiv Machine Learning

Jul 1

Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models

arXiv:2606. 31630v1 Announce Type: new Abstract: Language models increasingly write probabilistic programs (in NumPyro, Stan, or Pyro), but a program that compiles, runs, and passes every unit test can still be \emph{statistically} wrong -- a Gaussian likelihood for heavy-tailed data, a Poisson for over-dispersed counts, an invalid prior support, or a pathological parameterization.

By Jian Xu, Delu Zeng, John Paisley, Qibin Zhao

llms benchmarks

One email a morning, machine-written

Related stories

Toward Better Assessment of LLMs' Performance in Clinical Error Detection

Efficiency vs. Alignment: Investigating Safety and Fairness Risks in Parameter-Efficient Fine-Tuning of LLMs

Storage Is Not Strategy: State-Conditioned Support Control for LLM Unlearning

Constitutional Midtraining: Content Presence Drives Alignment Gains

Breaking Safety at the Token Boundary: How BPE Tokenization Creates Exploitable Gaps in LLM Alignment

Calibration, Not Compilation: Detecting and Repairing Misspecified Probabilistic Programs Written by Language Models