AI Paper Atlas

Research map / Large language models / Alignment and preferences

Mitigating Reward Hacking in LLM-based Recommendation: A Preference Optimization Approach

Heyu Chen, Junkang Wu, Guoqing Hu, Kexin Huang, Xiang Wang, Jiancan Wu

ICML 2026 · Large language models › Alignment and preferences · approach: DPO design choices

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
add SIRIUS pseudo-negativeSimPO w/o LN · LastFMHitRatio@1: 0.6633 -> 0.7315Improved
add SIRIUS pseudo-negativeSimPO w/o LN · GoodreadsHitRatio@1: 0.7010 -> 0.7857Improved
add SIRIUS pseudo-negativeSimPO w/o LN · SteamHitRatio@1: 0.8002 -> 0.8550Improved
add SIRIUS pseudo-negativeDPO · LastFMHitRatio@1: 0.6393 -> 0.6593Improved
add SIRIUS pseudo-negativeDPO · GoodreadsHitRatio@1: 0.6462 -> 0.6744Improved
add SIRIUS pseudo-negativeDPO · SteamHitRatio@1: 0.7333 -> 0.7757Improved
add SIRIUS to DPO instead of using S-DPO with three negativesDPO · GoodreadsHitRatio@1: 0.6744 -> 0.6744No clear effect
use DPO + SIRIUS instead of S-DPO with multiple negativesDPO · SteamHitRatio@1: 0.7712 -> 0.7757Improved
remove length normalization from SimPO (ablation)SimPO · LastFMHitRatio@1: 0.5928 -> 0.6633Improved
add SIRIUS to S-DPO-style training (SIRIUS runs keep improving past epoch 3)S-DPO · LastFM, Goodreads, SteamHitRatio@1Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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