Skipping the cross-encoder stage: LLM reranking alone costs 6x for +0.017 nDCG
CShorten30 · x · 2026-11-16
Tweet 9/17: skipping the cross-encoder and handing GPT-5.6 Luna hybrid retrieval's top-100 directly yields 0.375 nDCG@10 — barely above zerank-2's 0.358 on the same pool, at 6x the price. With zerank-2 first retaining top-100 from k=2,000, Luna's listwise pass reaches 0.482, showing the cheap cross-encoder prefilter is the pipeline's keystone.
Related event: Reranking Study Finds Two-Stage Pipeline Beats Fusion and Deeper Windows(13 posts)→
More from Research
- BeyondBinding challenge launches: AI-designed antibodies get wet-lab tested — QuanquanGu · 2026-11-17
- Singularity-Bench: Opus 5.5 with Claude Code beats human baselines on building AI systems — armand_ruiz · 2026-11-17
- Open Vision Banana: fully open recipe turns image generators into multi-task perception systems — NielsRogge · 2026-11-17
- WebArena-Pro debuts as largest browser-agent benchmark, top agents score 62.3% — xhluca · 2026-11-17
- NeurIPS 2026 paper: RL-trained skills transfer poorly to composed reasoning tasks — StanfordAILab · 2026-11-17
- OpenAI warns CoT monitorability will degrade as models scale, researchers push back — gleech · 2026-11-17