FOSTER: First-order Dataset Distillation for Text-based Sequential Recommendation

Collection of recent ML/IR papers (arXiv, 01 Jun 2026) covering dataset distillation for text-based sequential recommendation (FOSTER), representation effects in Retrieval-Augmented Generation (RAG) with a focus on answer retention, an inference-free multimodal sparse retriever for large-scale visual-document search (V-SPLADE), a typed claim network for citation stance and analytic tasks, dependency-aware credit assignment for LLM chain-of-thought in e-commerce relevance (Graph-GRPO), a data-centric reasoning compiler to reduce numerical hallucinations in financial QA (DCRC), multilingual IR (