Natural Language Processing Papers

Shop-R1: Rewarding LLMs to Simulate Human Behavior in Online Shopping via Reinforcement Learning. arxiv.org/abs/2507.17842

ylmmcl at Multilingual Text Detoxification 2025: Lexicon-Guided Detoxification and Classifier-Gated Rewriting. arxiv.org/abs/2507.18769

The Role of Orthographic Consistency in Multilingual Embedding Models for Text Classification in Arabic-Script Languages. arxiv.org/abs/2507.18762

Specification Self-Correction: Mitigating In-Context Reward Hacking Through Test-Time Refinement. arxiv.org/abs/2507.18742

MathOPEval: A Fine-grained Evaluation Benchmark for Visual Operations of MLLMs in Mathematical Reasoning. arxiv.org/abs/2507.18140

A New Pair of GloVes. arxiv.org/abs/2507.18103

Hybrid and Unitary Fine-Tuning of Large Language Models: Methods and Benchmarking under Resource Constraints. arxiv.org/abs/2507.18076

Privacy-Preserving Synthetic Review Generation with Diverse Writing Styles Using LLMs. arxiv.org/abs/2507.18055

Synthetic Data Generation for Phrase Break Prediction with Large Language Model. arxiv.org/abs/2507.18044

GrAInS: Gradient-based Attribution for Inference-Time Steering of LLMs and VLMs. arxiv.org/abs/2507.18043

NeuralDB: Scaling Knowledge Editing in LLMs to 100,000 Facts with Neural KV Database. arxiv.org/abs/2507.18028

Technical Report of TeleChat2, TeleChat2.5 and T1. arxiv.org/abs/2507.18013

Natural Language Processing for Tigrinya: Current State and Future Directions. arxiv.org/abs/2507.17974

Are LLM Belief Updates Consistent with Bayes' Theorem?. arxiv.org/abs/2507.17951

Evaluating the Performance of AI Text Detectors, Few-Shot and Chain-of-Thought Prompting Using DeepSeek Generated Text. arxiv.org/abs/2507.17944

One Whisper to Grade Them All. arxiv.org/abs/2507.17918

VeriMinder: Mitigating Analytical Vulnerabilities in NL2SQL. arxiv.org/abs/2507.17896

Dynamic and Generalizable Process Reward Modeling. arxiv.org/abs/2507.17849

Investigating Subjective Factors of Argument Strength: Storytelling, Emotions, and Hedging. arxiv.org/abs/2507.17409

Millions of $\text{GeAR}$-s: Extending GraphRAG to Millions of Documents. arxiv.org/abs/2507.17399