A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Nossa Iyamu
A deterministic, zero-model pipeline compiles passively captured screen activity into typed activity frames, reducing raw capture 86x and enabling 98.4% question-answering accuracy, while also measuring agent cost parameters.
Unknown
ThinKV is a thought-adaptive KV cache compression framework that exploits attention sparsity patterns to reduce memory usage in reasoning models.
Unknown
This paper systematically categorizes and explores KV cache compression techniques for transformer-based models, providing a structured analysis of methods to reduce memory overhead.
Unknown
This survey systematically reviews model quantization methods for computing-in-memory (CIM) accelerators, covering fundamental concepts and key techniques.
Unknown
This survey defines LLM agent memory as an externalized repository bridging frozen parametric knowledge and evolving environmental dynamics, and reviews its evolution.
Gail A. Carpenter, Boriana L. Milenova, Benjamin W. Noeske
Distributed ARTMAP combines MLP and ART advantages for fast, stable supervised learning with improved memory compression.
Unknown
This paper investigates whether multimodal large language models (MLLMs) can develop visual-spatial intelligence to remember and recall spaces from sequential visual observations, akin to human spatial memory.
Unknown
This paper introduces intrinsic memory agents, a heterogeneous multi-agent LLM system that uses structured contextual memory to manage conversation history and agent memory within token limits.
Unknown
Proposes DAM, a decision-theoretic framework for memory management in LLMs, treating memory as a utility-maximization problem rather than heuristic-based.
Unknown
Amory introduces an agentic reasoning framework that builds coherent narrative-driven memory for long-term conversational agents, addressing scalability by selectively processing relevant history.
Unknown
A-MemGuard is the first proactive defense framework for LLM agent memory, addressing security challenges in agent-based systems.
Unknown
MemReader introduces a reasoning-driven, selective memory extraction approach for long-term agent memory, reducing noise and improving performance.