🧪 Test?View on arXiv
Rethinking User Encoders: Routing Diffusion States for Scenario-Adaptive User Representation
Not provided in the abstract
multimodaluser representationdiffusion modelsrouting
2610.22400
Builder Relevance
1h ago80%
Abstract
The paper proposes a new framework, FINDERS, that enhances user representation by utilizing diffusion language models to overcome limitations of existing user encoders.
Reality Card
Core Claim
FINDERS outperforms state-of-the-art methods in both unified and scenario-adaptive settings by effectively utilizing a router-equipped diffusion language model.
Method / Result
Extensive experiments and online A/B tests demonstrate consistent performance improvements across ten industrial scenarios.
Limitations
The paper does not specify details on the reproducibility of the BehaviorQA corpus or the implementation of the FINDERS framework.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.
No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.