A Production-Ready RL Framework for Personalized Utility Tuning with Pareto Sweeping in Pinterest Recommender Systems
2026-05-19T08:52:15Z•2499a02752adb134d5adcc04bd4b890e6f181bcc60f328952b16eb53227a844c
LLM-adsad-auctionsbenchmarksbioinformaticscode-retrievaldeploymentevaluationfairnessindustrial-mljob-searchmulti-label-classificationpareto-frontierpersonalizationpopularity-biasproduction-systemsprotein-qarecommender-systemsreinforcement-learningretrieval-augmented-generationsoftware-engineering
What happened
A multi-paper arXiv feed (May 19, 2026) covering advances in recommender systems, retrieval, LLM applications, and benchmarks. Highlights include PRL-PUTS, a production-ready RL framework for personalized utility-weight tuning with inference-time Pareto sweeping (Pinterest); LARGER, a lexically anchored graph-based repository retrieval method for coding agents; an empirical study of Google AI Overviews’ effects on Reddit engagement; LERA, a two-stage LLM-enhanced ad-auction framework for chatbots; and several deployment-oriented retrieval/RAG contributions (RAPT for post-hoc thresholding, 2D-<
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_cs_ir
- Record identifier
- 2499a02752adb134d5adcc04bd4b890e6f181bcc60f328952b16eb53227a844c
- Enrichment time
- 2026-05-19T08:52:15Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.