Latest Articles for Language Models

Introducing models designed to improve natural language processing in Portuguese.

Learn how active prompt engineering improves tasks for language models.

This article reviews how chunk size affects AI-generated answers.

A fresh approach highlights surprising tokens to assess language model training data.

This study examines methods to enhance Italian language models in specialized fields.

A new method improves tamper resistance in open-weight language models.

Enhancing smaller language models like MiniCPM through effective fine-tuning practices.

Benchmark assesses large language models' ability to understand spatial relationships.

A new method analyzes language models by examining their specific characteristics.

This article examines how structured generation affects language model reasoning and comprehension.

OpenFactCheck provides a framework for evaluating the accuracy of language model outputs.

Innovative methods to enhance fairness in large language models.

A new method enhances synthetic data quality for better language model alignment.

A new system enhances speech recognition by using contextual keywords for better accuracy.

SAGE-RT creates synthetic data to improve language model safety assessments.

ArabLegalEval assesses LLMs' performance in handling Arabic legal information.

A new method to assess language model outputs using multiple LLM judges.

A new benchmark assesses language model agents for handling scientific data analysis.

New methods enhance small models' accuracy in telecommunications question answering.

ConflictBank offers insights into knowledge conflicts in language models.

This article explores the role of memorization in improving ICL performance.

Introducing a new model and benchmark for Russian text processing.

Researchers examine the reliability of metrics for language model safety.

A deep dive into how next-token prediction shapes language understanding in models.

FPDT offers a solution for training long-context LLMs more efficiently.

MemLong improves language models' ability to handle lengthy texts effectively.

This article analyzes how language models create realistic social networks and their biases.

This article discusses a new framework for enhancing reasoning in AI models.

Introducing a framework for generating creativity test items using language models.

A new method enhances long-text processing in language models for better answers.

LongGenBench assesses large language models in generating high-quality long text.

RAG remains vital in optimizing language model responses, especially with long texts.

This article assesses the effectiveness of sparse autoencoders in knowledge representation about cities.

A study on the impact of ICL and SFT on language model structure.

Study shows fine-tuning LLMs with TMs enhances translation quality for organizations.

This article discusses MLSAEs and their role in examining language model layers.

ECHO combines diverse reasoning patterns for better problem-solving in language models.

Study assesses language models on their ability to generate web application code.

AdaPPA enhances jailbreak attacks on language models by combining safe and harmful responses.

PF-PPO enhances language models by filtering out unreliable rewards for better code responses.