Yunzhi Lin — Blog https://yunzhilin.itreliable.com/blog/ Computer Systems Engineering undergraduate at the University of York specialising in AI. I build LLM data pipelines, hybrid retrieval engines and agent systems. en-gb Mon, 28 Sep 2026 00:00:00 GMT BM25 + Vector Search with RRF: how I stopped guessing at retrieval quality https://yunzhilin.itreliable.com/blog/hybrid-retrieval-rrf/ https://yunzhilin.itreliable.com/blog/hybrid-retrieval-rrf/ Why hybrid retrieval beats either retriever alone, how Reciprocal Rank Fusion quietly does the hard part, and why I now refuse to tune RAG without an evaluation harness. Mon, 28 Sep 2026 00:00:00 GMT RAG Retrieval Python What producing 50 training items a day taught me about LLMs https://yunzhilin.itreliable.com/blog/sft-rlhf-data-notes/ https://yunzhilin.itreliable.com/blog/sft-rlhf-data-notes/ Notes from my SFT/RLHF data internship at Zhipu: how training data is really made, the error taxonomy that saved my sanity, and why data quality is an engineering discipline. Wed, 12 Aug 2026 00:00:00 GMT LLM Data SFT/RLHF From O(n²) to 60 FPS: spatial hashing in a survivor game https://yunzhilin.itreliable.com/blog/spatial-hash-grid/ https://yunzhilin.itreliable.com/blog/spatial-hash-grid/ How a hash grid rescued Endless Onslaught from its own particle system, why profiling beats intuition, and what game-loop optimisation taught me about retrieval indexes. Tue, 30 Jun 2026 00:00:00 GMT TypeScript Game Dev Performance Designing a tool-calling agent that survives going offline https://yunzhilin.itreliable.com/blog/offline-capable-agent/ https://yunzhilin.itreliable.com/blog/offline-capable-agent/ Lessons from EchoList: why tool calling beats JSON generation, how to build a graceful fallback ladder, and the unglamorous details that make an AI feature feel trustworthy. Sat, 16 May 2026 00:00:00 GMT Agents Android LLM