https://www.solosoft.dev/tags/neurips/Recent content in NeurIPS on SoloSoftHugoen-usFri, 01 May 2026 00:00:00 +0000https://www.solosoft.dev/post/hipporag-memory-rag-2026/Fri, 01 May 2026 00:00:00 +0000https://www.solosoft.dev/post/hipporag-memory-rag-2026/<p>Retrieval-Augmented Generation (RAG) has become the standard approach for grounding LLM outputs in external knowledge. But standard RAG has a fundamental limitation: it treats each query independently, with no memory of past retrievals or ability to connect information across documents. <strong>HippoRAG</strong> takes inspiration from the human brain&rsquo;s hippocampus to overcome this, creating a long-term memory system that dramatically improves multi-hop question answering.</p> <p>Published at NeurIPS 2024 and available at <a href="https://github.com/OSU-NLP-Group/HippoRAG">github.com/OSU-NLP-Group/HippoRAG</a>, HippoRAG combines LLMs with knowledge graphs in a framework modeled on the hippocampal indexing theory of human memory. The result is a RAG system that builds a persistent knowledge structure from documents, enabling it to answer complex questions that require connecting information across multiple sources &ndash; achieving approximately 20% improvement over standard RAG on multi-hop QA benchmarks.</p>https://www.solosoft.dev/post/swe-agent-software-engineering-2026/Fri, 01 May 2026 00:00:00 +0000https://www.solosoft.dev/post/swe-agent-software-engineering-2026/<p>Princeton University&rsquo;s Natural Language Processing group has produced some of the most influential research in AI, and <strong>SWE-agent</strong> represents a landmark contribution to the emerging field of AI-driven software engineering. Rather than treating code generation as a stateless text completion problem, SWE-agent frames it as an interactive agent task: the model receives a GitHub issue, must explore the codebase to understand the context, formulate a fix, apply it, and verify the result.</p> <p>This approach mirrors how human developers actually work. When faced with a bug report, a developer does not immediately start writing code. They read the relevant files, search for related functions, check git history, run tests, and iteratively refine their understanding before making changes. SWE-agent replicates this workflow through a design innovation called the Agent-Computer Interface (ACI).</p>