15 papers
A Formal Kinetic Theory for Zeroth-Order Newton Dynamics:Stein-Corrected Hessian Estimation and Curvature--Variance Trade-offs
Shihao Ji, Mingyu Li, Zihui Song
Zeroth-order Newton-type methods are useful when gradients and Hessians are unavailable, but they behave quite differently from first-order gradient-free methods. We develop a kine…
Lagrange: An Open-Vocabulary, Energy-Based Sparse Framework for Generalized End-to-End Driving
Shihao Ji, HongXi Li, Zihui Song +1
Scaling end-to-end autonomous driving to complex, open-world environments requires perceptual models that generalize to anomalous scenarios and planners that produce kinematically…
Expected Value Alignment for Generative Reward Modeling in Formal Mathematics Verification
Shihao Ji, Haotao Tan, Zihui Song +1
Large Language Models (LLMs) are increasingly used with formal interactive theorem provers such as Lean 4. Scaling these systems with reinforcement learning or search methods requi…
Soft-NBCE: Entropy-Weighted Chunk Fusion for Long-Context
Shihao Ji, Mingyu Li, Zihui Song
The quadratic complexity of self-attention remains a bottleneck for Large Language Models (LLMs) processing ultra-long contexts. The Naive Bayes Cognitive Engine (NBCE) parallelize…
Data Scaling as Progressive Coverage of a Predictive Contribution Spectrum
Zihui Song, Shihao Ji, Hongxi Li +2
We investigate the hypothesis that real-data scaling laws are governed by progressive coverage of a latent predictive contribution spectrum rather than by token-frequency tails alo…
Xiaoice: Training-Free Video Understanding via Self-Supervised Spatio-Temporal Clustering of Semantic Features
Shihao Ji, Zihui Song
The remarkable zero-shot reasoning capabilities of large-scale Visual Language Models (VLMs) on static images have yet to be fully translated to the video domain. Conventional vide…
Credal Transformer: A Principled Approach for Quantifying and Mitigating Hallucinations in Large Language Models
Shihao Ji, Zihui Song, Jiajie Huang
Large Language Models (LLMs) hallucinate, generating factually incorrect yet confident assertions. We argue this stems from the Transformer's Softmax function, which creates "Artif…
L-MoE: End-to-End Training of a Lightweight Mixture of Low-Rank Adaptation Experts
Shihao Ji, Zihui Song
The Mixture of Experts (MoE) architecture enables the scaling of Large Language Models (LLMs) to trillions of parameters by activating a sparse subset of weights for each input, ma…
RAID-0e: A Resilient Striping Array Architecture for Balanced Performance and Availability
Yanzhao Jia, Zhaobo Wu, Zheyi Cao +3
This paper introduces a novel disk array architecture, designated RAID-0e (Resilient Striping Array), designed to superimpose a low-overhead fault tolerance layer upon traditional…
MyGO: Memory Yielding Generative Offline-consolidation for Lifelong Learning Systems
Shihao Ji, Zihui Song
Continual or Lifelong Learning aims to develop models capable of acquiring new knowledge from a sequence of tasks without catastrophically forgetting what has been learned before.…
OpenGrok: Enhancing SNS Data Processing with Distilled Knowledge and Mask-like Mechanisms
Lumen AI, Zaozhuang No. 28 Middle School, Shihao Ji +6
This report details Lumen Labs' novel approach to processing Social Networking Service (SNS) data. We leverage knowledge distillation, specifically a simple distillation method ins…
Enhancing Large Language Model Efficiencyvia Symbolic Compression: A Formal Approach Towards Interpretability
Lumen AI, Tengzhou No. 1 Middle School, Shihao Ji +6
Large language models (LLMs) face significant token efficiency bottlenecks in code generation and logical reasoning tasks, a challenge that directly impacts inference cost and mode…
Chinese Stock Prediction Based on a Multi-Modal Transformer Framework: Macro-Micro Information Fusion
Lumen AI, Tengzhou No. 1 Middle School, Shihao Ji +6
This paper proposes an innovative Multi-Modal Transformer framework (MMF-Trans) designed to significantly improve the prediction accuracy of the Chinese stock market by integrating…
Transformer^-1: Input-Adaptive Computation for Resource-Constrained Deployment
Lumen AI, Tengzhou No. 1 Middle School, Shihao Ji +6
Addressing the resource waste caused by fixed computation paradigms in deep learning models under dynamic scenarios, this paper proposes a Transformer architecture based on…
MyGO Multiplex CoT: A Method for Self-Reflection in Large Language Models via Double Chain of Thought Thinking
Shihao Ji, Zihui Song, Fucheng Zhong +4
Recent advancements in large language models (LLMs) have demonstrated their impressive abilities in various reasoning and decision-making tasks. However, the quality and coherence…