Natural Language Processing

Researchers from arXiv found that large language models share a common geometric structure in how context transforms concept representations, with displacement patterns that are semantically organized and transferable ac…

Researchers propose EmCom-Diffusion, a framework that measures visual reflection in emergent languages by reconstructing images from emergent messages using a finetuned diffusion model and comparing them to originals. Va…

A study in a graduate-level mobile robotics course found that a Socratic-Guidance (SG) tutor, which uses dialogic questioning, led to higher learning gains and more understanding-driven prompting strategies compared to a…

Researchers propose ASK+, an uncertainty-gated framework that improves small language model (SLM) guidance for reinforcement learning agents under partial observability. By providing trajectory-aware context and structur…

A study on GPT-5.2's Spanish-Chinese journalistic translations found that translation theory-driven prompts improved quality under expert evaluation, with the brief-oriented prompt scoring highest (MQM: 8.66 vs. 7.84), w…

Researchers propose TR-RAG, a teacher-regularized reinforcement learning method for cross-lingual retrieval-augmented generation that addresses language drift and unreliable evidence use. The method combines reward optim…

Researchers introduced CoC-Seduce, a benchmark for testing LLM rule adherence in semi-open textual environments, revealing that frontier models like GPT-5.4, Claude Sonnet 4.6, and Gemini 3.5 Flash are vulnerable to Rhet…

Researchers developed a spoken question answering system for Luxembourgish using text-to-speech augmentation, training a SLAM-style architecture with frozen Whisper and multilingual LLM backends. Multi-source synthetic t…

Researchers developed a multilingual, multimodal NLP framework for early detection of fake news and violence-driven mob activity, achieving 98% accuracy on a fused dataset of 138,256 Bangla and English samples. The syste…

Researchers propose Trajectory-Aware Commit Gating (TACG), a training-free decoder for diffusion language models that uses trajectory-aware signals to decide when to commit tokens, improving accuracy and efficiency on co…

Researchers propose Proactive Thinking, a framework enabling LLMs to pre-compute responses during conversational pauses, reducing latency. A training-free baseline using speculative continual thinking improves interactio…

Researchers propose Hierarchical Landmark Sparse (HiLS) Attention, a chunk-wise sparse attention mechanism that learns chunk selection end-to-end under the language-modeling loss. HiLS-Attention achieves performance comp…

Researchers found that large language models (LLMs) score up to 96% on CBT licensing exams but fail to apply CBT effectively in dialogue, defaulting to validation and reflection. A new metric, Protocol Leverage Force (F)…

Researchers found that induction heads in transformers implement two smoothing mechanisms—soft context-matching and BOS token pseudo-counts—that interpolate n-gram estimates, enabling in-context learning to regularize ra…

Researchers formalized medical diagnosis as an iterative evidence-seeking task and used reinforcement learning with verifiable rewards to train large language models to autonomously gather clinical evidence. Their framew…

Researchers propose MPSelectTune, a two-stage fine-tuning method that improves concept unlearning in large language models by focusing on the prompt type with the highest concept accuracy. The approach achieves 2-15% mai…

Researchers propose KARMA, a method that uses knowledge graphs to generate slot-aligned contrastive candidates for preference optimization in large language models, addressing the Resolution Mismatch Problem. KARMA outpe…

Researchers introduced APeB, a benchmark for evaluating personalization in LLM-powered agents, finding that models struggle with raw, underspecified queries due to ineffective use of interaction histories. A proposed pip…

Researchers propose a reinforcement learning with verifiable rewards (RLVR) method for data-efficient adaptation of audio-language models to code-switched automatic speech recognition (ASR). Using Qwen2-Audio across 10 l…

Researchers have developed a method to distinguish AI-generated fake speech from natural human speech by analyzing vowel spectral distributions using the Wasserstein metric and persistent homology, with Japanese as a tes…