LinkedIn's TransX report: sequence transduction for recsys, +6.02% CTR in A/B tests
_reachsumit · x · 2026-10-16
LinkedIn presents TransX, a sequence-transduction framework with co-designed infrastructure for industrial recommendation. It splits sequence encoding, behavior memory, and candidate-conditioned fusion into independently scalable components; training amortizes causal trajectory computation across behavioral prefixes, and serving disaggregates bulk prefill, nearline refresh, and online prediction via compact behavior-memory KV so models scale without raising serving cost. Findings include anisotropic scaling returns (encoder scaling saturates beyond moderate depth; more supervision shifts allocation toward candidate fusion). Production A/B tests: core model improves CTR by 6.02% and conversion by 4.45%; post-trained retrieval and preference-ranking variants add CTR +2.96% and conversion +1.33%.
More from Infra
- RL train-sample systems are frontier AI's hardest infra problem — and a trade secret — brianryhuang · 2026-10-16
- Local tuning of GLM-5.3-Flash on 2×GB10 lifts decode from 49 to 58 tok/s — EAccelerate_42 · 2026-10-16
- 3x DGX Spark cluster runs GLM-5.3-Flash locally behind a proxy, 1.75M tokens/day at 119W — Teknium · 2026-10-16
- Optuna creator Akiba: when 2 hyperparameters cost $20M, what's left for BO/HPO — Stefania_druga · 2026-10-16
- Amazon Proposes Per-Request Validated Evidence Caching for Production RAG — _reachsumit · 2026-10-16
- llama.cpp PR boosts CPU inference speed with merged AVX2 tiled accumulators — jacek2023 · 2026-10-16