AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Temporal Preference Optimization for Unsupervised Retrieval

HyunJin Kim, Jaejun Shim, Young Jin Kim, JinYeong Bak

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Recommendation representation learning · 0 citations (OpenAlex)

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
train with TRPO preference loss on temporally aligned vs misaligned documentsContriever · SituatedQA 2018 explicitnDCG@5: 29.30 -> 43.93Improved
train with TRPO preference lossContriever · SituatedQA 2018 implicitnDCG@5: 29.89 -> 44.11Improved
use TRPO preference loss instead of temporal contrastive loss (ablation)Contriever · SituatedQA 2018 explicitnDCG@5: 35.00 -> 43.93Improved
TPOUR Contriever versus 72.7x larger Qwen3-Embedding-8B with time-aware instructionTPOUR Contriever · SituatedQA 2021 explicitnDCG@5: 38.51 -> 40.21Improved
add query rewriting to Qwen3-Embedding-8B (ablation)Qwen3-Embedding-8B · SituatedQA explicit 2018 vs 2021nDCG@5: 2018: 30.45 -> 40.92; 2021: 35.17 -> 34.46Mixed
add time-aware instruction to Qwen3-Embedding-8B (ablation)Qwen3-Embedding-8B · SituatedQA explicitnDCG@5: 2018: 30.45 -> 32.69; 2021: 35.17 -> 38.51Improved
interpolate time vectors at alpha 0.5 between 2018 and 2021 retrievers (ablation)TPOUR Contriever · SituatedQAnDCG@5: 29.05 / 31.72 -> 35.71Improved
use best interpolation alpha instead of directly fine-tuned retriever (ablation)TPOUR Contriever · RealTimeQAnDCG@5: 37.30 -> 38.77Improved
add time-vector extrapolation from 2018/2021/2022 vectorsTPOUR Contriever · RealTimeQA DecembernDCG@5: 27.78 -> 30.00Improved
use mixture of TPOUR encoders for timestamp predictionTPOUR · Year classificationaccuracy (%): 50.18 -> 76.56Improved
use 10 encoders instead of 2 in mixture-of-TPOUR (ablation)Mixture-of-TPOUR · Year classificationaccuracy (%): 74.56 -> 76.56Improved
set loss weight lambda to moderate values 0.7-0.85 (ablation)TPOUR · SituatedQA/RealTimeQAnDCGImproved

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