Memory Tax quantifies the net value of agent memory systems
Original titleMemory Tax: Dissecting the Net Value of Agent Memory
- Outlet
- Hugging Face
- Published
- Importance
- 3/5
- Trending
- 16
Summary
Through large-scale evaluation across multiple benchmarks, the authors find many dedicated memory systems underperform naive RAG, indicating negative net value. They introduce the concept of 'memory tax' to measure losses from memory systems alongside gains, shifting focus from standalone performance to net value.
Also reported by 8 outlets
- Takara TLDRRetrieval as Traversal: Value-Indexed Memory for Stateful Long-Context Agents
- Hugging FaceHABIT-Bench: Evaluating Habit Induction from Longitudinal Weak Evidence in Agent Memory
- Takara TLDRPreserving Utility in Continual LLM Unlearning via Orthogonal Projections
- Takara TLDRGlance: The Generator Is Its Own Reranker
- Takara TLDRWizard: Manipulate Any World in Real Time via Agentic Control
- Takara TLDRRFW-A: Agentic Procedural Generation of 400K Articulated Assets over 500 Categories
- Takara TLDRWorldProbe: When Video World Models Fall Apart under Long-Horizon Spatial Exploration
- Takara TLDRDistilling Large-Scale Machine Learning Interatomic Potentials to Predict Long-Timescale Electrolyte Transport Properties
Original excerpt
Memory is increasingly used in long-horizon agents, yet the net value of building memory over direct retrieval from raw history remains largely unclear. Through a large-scale evaluation across multiple benchmarks and model backbones, we find many dedicated memory systems underperform naive RAG in many cases, indicating a negative net value. Taking RAG as a reference, we dissect the net value of memory into the gains and losses over direct retrieval, termed memory benefit and memory tax. These gains and losses arise because memory process changes how the raw information is organized and accessed. We view such changes as the double-edged deviations from the raw history. They are originally designed and introduced to benefit the task, for example by reorganizing facts or distilling preferences, but some may also remove or obscure evidence, giving rise to memory tax. Since memory benefit reflects the expected gains of its design, we then focus more on the overlooked memory tax, study how it arises from the system and evaluation side, and generalize the concept to more application scenarios of memory. We hope the notion of memory tax can provide a new lens and shift more attention from standalone performance to the net value of memory, informing more reliable memory designs and evaluations. We will open-source the code and results to the community. Project Page: https://hexuan21.github.io/MemoryTax/