AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Improving LLM-Based Recommenders with Conservative Generative Flow Networks

Xuan Yu, Feng Niu, Rui Zhu, Yudong Zhang, Xu Wang, Yang Wang

ICML 2026 · Reinforcement learning › RL from feedback · approach: RL fine-tuning of generative models

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Flower (SubTB) with conservative SubTB (CFlower)LLM recommender · Amazon Video GamesNDCG@5: 0.0546 -> 0.0646Improved
replace Flower with CFlowerLLM recommender · Amazon Movies and TVNDCG@5: 0.0956 -> 0.1036Improved
replace Flower with CFlowerLLM recommender · Amazon CDs and VinylNDCG@5: 0.0702 -> 0.0737Improved
replace Flower with CFlowerLLM recommender · Amazon Video GamesHR@5: 0.0803 -> 0.0841Improved
replace Flower with CFlowerLLM recommender · Amazon CDs and Vinylexposure entropy H: 7.928 -> 8.132Improved
CFlower versus D3 (accuracy-oriented baseline)CFlower · Amazon CDs and VinylNDCG@5: 0.0810 vs 0.0737Hurt
replace Flower with CFlower for distribution matchingLLM recommender · target item distributionJaccard similarity: 0.799 -> 0.854Improved
replace Flower with CFlower for distribution matchingLLM recommender · target item distributiontotal variation distance: 0.436 -> 0.332Improved
use CFlower instead of BIGRec as reference policy for S-DPOS-DPO · Amazon CDs and VinylNDCG@5: 0.0716 -> 0.0805Improved
use CFlower instead of Flower as reference policy for S-DPOS-DPO · Amazon Video GamesNDCG@5: 0.0640 -> 0.0737Improved
use CFlower instead of Flower as reference policy for DMPODMPO · Amazon Movies and TVNDCG@5: 0.0979 -> 0.1002Hurt
add offline constraints (dataset-constrained learning) (ablation)SubTB/CSubTB · Amazon recommendationNDCG/HRImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy