NSGR is a tree-structured generative reranker that progressively generates optimal lists via next-scale expansion and multi-scale neighbor loss to balance perspectives and align training signals.
Title resolution pending
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
method 1
citation-polarity summary
representative citing papers
Graph Retention Networks extend retention to dynamic graphs to enable parallelizable training, O(1) inference, and chunkwise long-term training while delivering competitive performance with major efficiency gains.
HALo uses smartglasses IMU head orientation to localize conversation partners' acoustic zones, achieving 21% better performance with known partner count, while CoCo classifies partner numbers at 0.74 accuracy using only IMU data.
FLAME condenses ensemble diversity into a single network via modular ensemble simulation and guided mutual learning during training, delivering ensemble-level performance with single-network inference speed on sequential recommendation tasks.
A tempo-relational neural architecture jointly models temporal and relational aspects of team interactions to outperform prior approaches on team performance prediction and enable efficient multi-task prediction of team constructs.
AlignedServe uses prefix-aware batching, large CPU in-flight request pools, batch scheduling, and GPU-to-GPU KV prefetching to raise decoding throughput up to 1.98x and cut latency up to 7.4x versus prior serving systems.
CDGLT achieves SOTA on MET-Meme for multimodal metaphor identification by using SLERP-based concept drift and prompt-adapted LayerNorm tuning with reduced compute.
CCL-D detects slow/hang anomalies in CCL for distributed training via lightweight tracing probes and an intelligent analyzer, achieving near-complete coverage and 6-minute rank localization on a 4000-GPU cluster over one year.
citing papers explorer
-
Next-Scale Generative Reranking: A Tree-based Generative Rerank Method at Meituan
NSGR is a tree-structured generative reranker that progressively generates optimal lists via next-scale expansion and multi-scale neighbor loss to balance perspectives and align training signals.
-
Graph Retention Networks for Dynamic Graphs
Graph Retention Networks extend retention to dynamic graphs to enable parallelizable training, O(1) inference, and chunkwise long-term training while delivering competitive performance with major efficiency gains.
-
Towards Localizing Conversation Partners using Head Motion
HALo uses smartglasses IMU head orientation to localize conversation partners' acoustic zones, achieving 21% better performance with known partner count, while CoCo classifies partner numbers at 0.74 accuracy using only IMU data.
-
FLAME: Condensing Ensemble Diversity into a Single Network for Efficient Sequential Recommendation
FLAME condenses ensemble diversity into a single network via modular ensemble simulation and guided mutual learning during training, delivering ensemble-level performance with single-network inference speed on sequential recommendation tasks.
-
Boosting Team Modeling through Tempo-Relational Representation Learning
A tempo-relational neural architecture jointly models temporal and relational aspects of team interactions to outperform prior approaches on team performance prediction and enable efficient multi-task prediction of team constructs.
-
AlignedServe: Orchestrating Prefix-aware Batching to Build a High-throughput and Computing-efficient LLM Serving System
AlignedServe uses prefix-aware batching, large CPU in-flight request pools, batch scheduling, and GPU-to-GPU KV prefetching to raise decoding throughput up to 1.98x and cut latency up to 7.4x versus prior serving systems.
-
Concept Drift Guided LayerNorm Tuning for Efficient Multimodal Metaphor Identification
CDGLT achieves SOTA on MET-Meme for multimodal metaphor identification by using SLERP-based concept drift and prompt-adapted LayerNorm tuning with reduced compute.
-
CCL-D: A High-Precision Diagnostic System for Slow and Hang Anomalies in Large-Scale Model Training
CCL-D detects slow/hang anomalies in CCL for distributed training via lightweight tracing probes and an intelligent analyzer, achieving near-complete coverage and 6-minute rank localization on a 4000-GPU cluster over one year.