Search papers, labs, and topics across Lattice.
Affiliation:
6
0
7
40
Token-level credit assignment can drastically improve the effectiveness of generative document retrieval, leading to better alignment between generation and relevance.
Agents can now seamlessly transition between seeking and following tasks in dynamic environments, setting a new benchmark for embodied AI performance.
Malicious knowledge injection can be executed effectively through parameter manipulation in RAG systems, exposing a significant security vulnerability.
DeepRubric achieves comparable performance to state-of-the-art models while slashing reinforcement learning resource consumption by 13 times.
Generative retrieval gets a reasoning upgrade: ThinkGR interleaves chain-of-thought with docid generation, boosting multi-hop retrieval performance by 6.86%.
You can boost LLM-based recommendations across completely disjoint domains without sacrificing user privacy by using semantic bridges and federated learning.