Imagen de Portada

WMF-AM: Probing LLM Working Memory via Depth-Parameterized Cumulative State Tracking

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Hou, Dengzhe, Jiang, Lingyu, Li, Deng, Li, Zirui, Lin, Fangzhou, Yamada, Kazunori D
Formato:	Preprint
Publicado:	2026
Materias:	Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Same Brain, Different Prediction: How Preprocessing Choices Undermine EEG Decoding Reliability
por: Hou, Dengzhe, et al.
Publicado: (2026)

PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning
por: Jiang, Lingyu, et al.
Publicado: (2026)

Physics-Aware Video Instance Removal Benchmark
por: Li, Zirui, et al.
Publicado: (2026)

Learning Random Numbers to Realize Appendable Memory System for Artificial Intelligence to Acquire New Knowledge after Deployment
por: Yamada, Kazunori D
Publicado: (2024)

Position: Human-Centric AI Requires a Minimum Viable Level of Human Understanding
por: Lin, Fangzhou, et al.
Publicado: (2026)

Constitutive Components for Human-Like Autonomous Artificial Intelligence
por: Yamada, Kazunori D
Publicado: (2025)

KANMixer: a minimal KAN-centered mixer for long-term time series forecasting
por: Jiang, Lingyu, et al.
Publicado: (2025)

AdaptFuse: Training-Free Sequential Preference Learning via Externalized Bayesian Inference
por: Lin, Fangzhou, et al.
Publicado: (2026)

CAPS: Cascaded Adaptive Pairwise Selection for Efficient Parallel Reasoning
por: Lin, Fangzhou, et al.
Publicado: (2026)

TimePre: Bridging Accuracy, Efficiency, and Stability in Probabilistic Time-Series Forecasting
por: Jiang, Lingyu, et al.
Publicado: (2025)

EcphoryRAG: Re-Imagining Knowledge-Graph RAG via Human Associative Memory
por: Liao, Zirui
Publicado: (2025)

Procedural Content Generation via Generative Artificial Intelligence
por: Mao, Xinyu, et al.
Publicado: (2024)

Capturing LLM Capabilities via Evidence-Calibrated Query Clustering
por: Wu, Fangzhou, et al.
Publicado: (2026)

Explore Briefly, Then Decide: Mitigating LLM Overthinking via Cumulative Entropy Regulation
por: Bin, Yi, et al.
Publicado: (2025)

CV-Arena: An Open Benchmark for Instructional Computer Vision Problem Solving with Human-AI Collaborative Preferences
por: Lin, Fangzhou, et al.
Publicado: (2026)

How Memory Management Impacts LLM Agents: An Empirical Study of Experience-Following Behavior
por: Xiong, Zidi, et al.
Publicado: (2025)

A Memory-Augmented LLM-Driven Method for Autonomous Merging of 3D Printing Work Orders
por: Liu, Yuhao, et al.
Publicado: (2025)

Diff-MN: Diffusion Parameterized MoE-NCDE for Continuous Time Series Generation with Irregular Observations
por: Zhang, Xu, et al.
Publicado: (2026)

FOAM: Blocked State Folding for Memory-Efficient LLM Training
por: Wen, Ziqing, et al.
Publicado: (2025)

Depth Gives a False Sense of Privacy: LLM Internal States Inversion
por: Dong, Tian, et al.
Publicado: (2025)

Rethinking Importance Sampling in LLM Policy Optimization: A Cumulative Token Perspective
por: Zhang, Yuheng, et al.
Publicado: (2026)

PRISM: Probing Reasoning, Instruction, and Source Memory in LLM Hallucinations
por: Wu, Yuhe, et al.
Publicado: (2026)

DrugPilot: LLM-based Parameterized Reasoning Agent for Drug Discovery
por: Li, Kun, et al.
Publicado: (2025)

Efficient Training-Free Online Routing for High-Volume Multi-LLM Serving
por: Wu, Fangzhou, et al.
Publicado: (2025)

Influencing LLM Multi-Agent Dialogue via Policy-Parameterized Prompts
por: Bo, Hongbo, et al.
Publicado: (2026)

Implementing Cumulative Functions with Generalized Cumulative Constraints
por: Schaus, Pierre, et al.
Publicado: (2025)

Loss Distillation via Gradient Matching for Point Cloud Completion with Weighted Chamfer Distance
por: Lin, Fangzhou, et al.
Publicado: (2024)

Beyond Dialogue Time: Temporal Semantic Memory for Personalized LLM Agents
por: Su, Miao, et al.
Publicado: (2026)

Accelerate Scaling of LLM Finetuning via Quantifying the Coverage and Depth of Instruction Set
por: Wu, Chengwei, et al.
Publicado: (2025)

MemoryKT: An Integrative Memory-and-Forgetting Method for Knowledge Tracing
por: Lin, Mingrong, et al.
Publicado: (2025)

HAGE: Harnessing Agentic Memory via RL-Driven Weighted Graph Evolution
por: Jiang, Dongming, et al.
Publicado: (2026)

Incompressible Knowledge Probes: Estimating Black-Box LLM Parameter Counts via Factual Capacity
por: Li, Bojie
Publicado: (2026)

MemLineage: Lineage-Guided Enforcement for LLM Agent Memory
por: Ouyang, Ciyan, et al.
Publicado: (2026)

Parameterizing Dataset Distillation via Gaussian Splatting
por: Jiang, Chenyang, et al.
Publicado: (2025)

Decomposing LLM Self-Correction: The Accuracy-Correction Paradox and Error Depth Hypothesis
por: Li, Yin
Publicado: (2025)

LLM-Assisted Semantic Alignment and Integration in Collaborative Model-Based Systems Engineering Using SysML v2
por: Li, Zirui, et al.
Publicado: (2025)

MemR$^3$: Memory Retrieval via Reflective Reasoning for LLM Agents
por: Du, Xingbo, et al.
Publicado: (2025)

MetaState: Persistent Working Memory Enhances Reasoning in Discrete Diffusion Language Models
por: Xia, Kejing, et al.
Publicado: (2026)

Eco-driving for Electric Connected Vehicles at Signalized Intersections: A Parameterized Reinforcement Learning approach
por: Jiang, Xia, et al.
Publicado: (2022)

Empowering Working Memory for Large Language Model Agents
por: Guo, Jing, et al.
Publicado: (2023)