MICE paper pushes cross-encoders from re-ranking to end-to-end retrieval
antoine_chaffin · x · 2026-10-27
An EMNLP 2026 paper introduces MICE, pushing the cross-encoder Pareto frontier and turning theoretical pre-computation gains into an operational end-to-end retrieval architecture rather than a re-ranking-only component. Retrieval researcher antoinechaffin endorses richer interactions at scale, arguing simple dot products should go even if the replacement isn't MaxSim, and notes open releases finally enable apples-to-apples comparisons beyond MSMARCO-only training mixes.
More from Research
- Open Models Match GPT-6 on MedMCQA but Trail 17 Points on Hard Clinical Cases — keunwoochoi · 2026-10-27
- Health-Focused Benchmark Reranks LLMs: Muse Spark Open Jumps from #9 to #1 — keunwoochoi · 2026-10-27
- Open-source Latch compiles PyTorch into training megakernels, 2.75x faster on H100 — sguada · 2026-10-27
- Goodfire commits $10M to a 'Human Genome Project for AI' to reverse-engineer frontier models — CatAstro_Piyush · 2026-10-27
- π0.5 study: raising demos from 10 to 438 lifts Place Ball in Bowl success from 7% to 83% — stingning · 2026-10-27
- HandsOn benchmark: 66 verified tasks and 18,000 human demos for dexterous manipulation — stingning · 2026-10-27