Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Researchers propose COOPA, a modular LLM agent architecture for operations research problems that combines iterative confidence-based modeling, element-level provenance explanations, and multi-solver routing. In benchmar…
Researchers introduced an axiomatic evaluation framework for latent thought representations in LLMs, comprising four axioms (Causality, Minimality, Separability, Stability) that measure representation quality independent…
Team HSA_CORAL submitted to the FinCausal 2026 shared task on extracting cause-effect relations from financial narratives via extractive question answering in English and Spanish. Their best system, GPT-4.1 Mini fine-tun…
Researchers used a developmental approach to study how neural language models learn statistical patterns, finding that Transformers first acquire abstract global statistics before local dependencies, with early over-gene…
Researchers at arXiv identified a memory-update gap in LLM agents, where accuracy drops from 92% to 77% when using bounded memory instead of full context, even with frontier models like gpt-5.4. The gap persists across m…
Researchers propose the world wide AI-model network (AI-ModelNet), a paradigm for interconnecting heterogeneous AI models to enable capability sharing and collaborative reasoning, inspired by the Internet's development. …
A developer argues that philosophy is crucial for understanding AI and software engineering, using the analogy that a gun cannot read a bedtime story to illustrate that tools have inherent limitations. The developer cont…
Researchers introduced the context-ready transformer, a recurrent neural network architecture that pre-contextualizes tokens before they enter a transformer block, achieving faster inference and competitive performance. …
Researchers introduced Yuvion LLM, a large language model designed for adversarially robust content and AI safety, addressing safety failures from strategic attacks. The model, which outperforms larger baselines like GPT…
Researchers introduced DiscoBench, a benchmark for clarification-aware deep search, to evaluate whether LLM-powered search agents can identify ambiguity, ask effective questions, and recover correct reasoning paths. The …
Researchers propose a protocol to mitigate p-hacking in LLM-based research by preregistering experiments and running them on the first eligible model released after registration. Across 20 models and 11 configurations, t…
Researchers propose a three-stage training paradigm to internalize future-aware planning in LLM agents, enabling them to simulate outcomes before acting. The approach, evaluated on search and math reasoning tasks, outper…
Researchers introduced LPES, a layer-specific positional embedding scaling method that mitigates the 'lost-in-the-middle' problem in LLMs by assigning distinct scaling factors to each layer, achieving up to 11.2% accurac…
Researchers at arXiv introduced a persona-driven rewriting pipeline that conditions user turns on low agreeableness to reduce jailbreak susceptibility and harmful output rates in LLMs fine-tuned for social warmth, preser…
Researchers propose a symbolic feedback-driven iterative self-refinement framework to improve the robustness and reliability of large language models in long-horizon planning tasks. The framework uses natural language pr…
Researchers propose HybridCodec, a novel approach combining discrete tokens with continuous residuals to improve speech representation in language models, reducing information loss and autoregressive steps while retainin…
Researchers found that personality prompting of LLMs affects multi-agent team performance only in open-ended and competitive tasks, not in structured coding. Low agreeableness degraded collaboration and bargaining outcom…
Researchers proposed DysLexLens, a low-resource LLM framework to analyze dyslexic learners' experiences with AI tools using Reddit forum data. The framework filters noisy posts, integrates knowledge-graph reasoning, and …
Researchers found that explicit reasoning does not always improve multimodal emotion recognition accuracy, with fast thinking often outperforming slow thinking. They propose MER-R1, a reinforcement learning framework tha…
Researchers propose Tree of Evidence (ToE), a hierarchical claim verification framework that uses reinforcement learning-driven multi-source retrieval and argument tree aggregation to combat fake news and AI-generated mi…