-
Steer2Adapt: Dynamically Composing Steering Vectors Elicits Efficient Adaptation of LLMs
score 2
关键词(5): efficient, lightweight, inference, reasoning, safety
-
Cross-View World Models
score 2
关键词(3): agent, agents, planning
-
VertCoHiRF: Decentralized Vertical Clustering Beyond k-means
score 2
关键词(2): agent, agents
-
Fin-RATE: A Real-world Financial Analytics and Tracking Evaluation Benchmark for LLMs on SEC Filings
score 2
关键词(6): deployment, retrieval-augmented, reasoning, benchmark, evaluation
-
Progressive Searching for Retrieval in RAG
score 2
关键词(4): efficient, RAG, search, cost
-
Principled Synthetic Data Enables the First Scaling Laws for LLMs in Recommendation
score 2
关键词(5): scaling, preference, pre-training, recommendation, synthetic data
-
KRONE: Hierarchical and Modular Log Anomaly Detection
score 2
关键词(2): efficient, efficiency
-
Parallel Track Transformers: Enabling Fast GPU Inference with Reduced Synchronization
score 2
关键词(8): efficient, efficiency, fast, serving, inference
-
Semantic Search At LinkedIn
score 2
关键词(10): scaling, efficiency, compression, distillation, pruning
-
LUCID-SAE: Learning Unified Vision-Language Sparse Codes for Interpretable Concept Discovery
score 2
关键词(4): alignment, multimodal, vision-language, evaluation
-
Beyond Accuracy: Risk-Sensitive Evaluation of Hallucinated Medical Advice
score 2
关键词(2): evaluation, safety
-
Action-to-Action Flow Matching
score 2
关键词(7): efficiency, fast, inference, latency, real-time
-
High Fidelity Textual User Representation over Heterogeneous Sources via Reinforcement Learning
score 2
关键词(2): latency, search
-
Intent Mismatch Causes LLMs to Get Lost in Multi-Turn Conversation
score 2
关键词(2): scaling, alignment
-
RAPiD: Real-time Deterministic Trajectory Planning via Diffusion Behavior Priors for Safe and Efficient Autonomous Driving
score 2
关键词(8): efficient, deployment, real-time, diffusion, planning