Natural Language Processing

Researchers have developed Guidance Contrastive Policy Optimization (GCPO), a new algorithm that assigns per-token credit by contrasting model predictions under positive and negative prompts, addressing the limitation of…

A new study analyzing 17 language models from 410 million to over 100 billion parameters found that instruction-tuned systems systematically collapse linguistic entropy along discourse and structural dimensions, with mea…

Researchers have introduced Micro-Macro Retrieval (M2R), a new framework designed to reduce hallucination in large language models during long-form text generation. The system addresses the problem of factual errors by e…

Researchers have introduced Thoughts-as-Planning, a framework that formalizes reasoning chain optimization in large language models as a sequential decision-making process over a latent semantic space. The method models …

A new study from arXiv (2605.29027) found that tonal variations in prompts cause systematic but model-dependent accuracy shifts in large language models (LLMs) on objective multiple-choice questions. Testing four cost-ef…

Researchers introduced GPF-LiveNews, a streaming evaluation protocol that audits how large language models frame breaking news for different audience groups. The protocol expands BBC and Reuters news anchors across 42 id…

A new paired-minimum detectable effect (MDE) budget for 4-bit quantization benchmarks, derived from classical sample-size calculations, provides benchmark designers a one-line formula to pre-register reliability claims b…

A new study from researchers has systematically measured behavioral consistency in multi-step tool-calling LLM agents, finding that the same agent often fails to select the same tools, in the same order, or with the same…

Researchers have developed SERC, a semantic error correction framework inspired by LDPC codes that treats LLM text generation as a noisy communication channel to detect and fix hallucinations. The training-free, model-ag…

Researchers have developed NRLB (No Reader Left Behind), a multi-agent framework for plain language summarization that simulates three reader groups: elementary students, non-native speakers, and readers with attention d…

A new study from the University of Hamburg found that LLM-generated peer reviews for scientific papers show only limited alignment with human evaluations, with alignment varying significantly across different prompts and…

Frontier large language model-based agents from Anthropic and OpenAI matched the performance of trained human biocurators in linking free-text phenotype descriptions to ontology terms, a task known as phenotype annotatio…

Researchers conducted the first comprehensive comparison of Dutch syllabification algorithms, finding that data-driven approaches outperformed knowledge-based methods across most datasets. A newly developed deep learning…

A new study systematically examined the effect of transformer-based language model size on topic quality, finding that model size had a negligible impact on coherence and divergence metrics. Researchers tested seven mode…

Researchers have developed S3MEM, a structured memory framework that improves long-horizon interactive question answering by converting agent trajectories into query-aligned evidence. In tests across four environments, S…

RightNowAI released RightNow-Arabic-0.5B-Turbo, a 518M-parameter Arabic-specialized language model that outperforms all existing sub-1B open models on Arabic benchmarks while running on edge devices. The model, built by …

Researchers have developed a method to detect and track concepts within large language models (LLMs) by creating datasets that delineate specific ideas, training linear probes to identify those concepts across model laye…

Researchers have developed COLAGUARD, a new safety guardrail for large language models that transfers multi-step reasoning into a continuous latent space to reduce latency. In tests across ten moderation settings and eig…

A developer used Claude Code to build an offline meeting transcriber that processes audio into Granola-style markdown notes entirely on a MacBook, ensuring sensitive meeting recordings never leave the device. The tool ru…

A RAG system's output-stage PII masking fails to prevent data exfiltration because the LLM has already received confidential context before the filter runs, enabling three unstoppable leak classes: creative paraphrasing,…