// TAG: RESEARCH

This paper introduces 'method-level change-proneness' as a novel metric to improve black-box test suite minimization, making testing more efficient …

This explainer details the Pegasus spyware attack on European Parliament members, covering technical compromise, mobile security implications, and …

Explaining the new SPEC CPU 2026 benchmark suite, its modernization efforts, and its impact on CPU, memory, and compiler development for the next …

This paper argues that DNS names, not IP addresses, are now the fundamental identifiers for service identity, reachability, load balancing, and trust …

Labyrinth 1.1 introduces a new sub-protocol to ensure end-to-end encrypted messages survive device loss, switching, and long sign-in gaps, making E2EE …

TeamTR introduces trust-region fine-tuning to prevent shared-context drift, a critical failure mode in multi-agent LLM systems, significantly …

This explainer clarifies recent LLM benchmark results, addressing claims of 0% scores and detailing actual performance on complex software engineering …

This paper reveals that instruction-tuned LLMs can produce fair outputs while still retaining causally potent and asymmetrically biased internal …

This paper proposes 'face density' as a novel metric to quantify data complexity, particularly for instance counting tasks in computer vision.

Mistral AI's Vox-Trainer is a new multimodal model capable of understanding and generating both spoken audio and text, with accessible fine-tuning …

This paper introduces an actor-verifier AI architecture that enhances reliability and interpretability in safety-critical systems by having a primary …

This paper introduces a novel method to train LLMs to internally recognize their own hallucinations by distilling weak, external hallucination signals …

RAGEN-2 identifies and measures 'reasoning collapse' in multi-turn LLM agents, where internal thought processes degrade despite initial task success, …

SymptomWise proposes a framework that enhances AI reliability and interpretability by separating natural language understanding (handled by an LLM) …

Google's TurboQuant algorithm slashes LLM KV cache memory by 6x and delivers up to 8x attention speedup with zero accuracy loss, significantly …

MTA-Agent introduces a modular, multi-turn agent framework that enhances Multimodal Large Language Models (MLLMs) by integrating specialized tools for …