Memory and context
41 notes, newest first. Includes the vault tags memory, context management, memory management, memory systems, agent memory, long-context.
Use and Disuse: Intent-Structured Experience Consolidation for Memory and Learning in LLM Agents
The preprint proposes Hippocam, a hierarchical memory and continual-learning architecture that organizes ongoing work as nested intents. Active context stays centered on the current intent, while completed intents are…
Causal Memory Policy: Making Memory Utility Identifiable by Intervening on Retrieval
The preprint identifies a retrieval-level positivity problem: memories that are never retrieved cannot have their utility identified by store-level interventions. It proposes Causal Memory Policy, which reserves…
VISTA: A Visual Harness for Reasoning in an Interactive World
The technical report introduces a visual harness that gives a multimodal model long-horizon visual context through direct perception, lossless visual memory, active retrieval, and input reorganization. The abstract…
Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning
The preprint introduces an inference-time harness in which a controller explicitly decides what partial work to reuse, whether to restart, when to stop, and which next work to dispatch. Workers perform task-level…
Share-Borne AI Virus: Memory-Hopping Attacks Across LLM Agents
The preprint describes artifact-mediated propagation, in which adversarial content inserted into a shared artifact is stored in one assistant's persistent memory, reproduced in a later artifact, and acquired by another…
Mutable Transcripts: Mitigating Context Pollution through Editable Conversation State
The preprint proposes allowing users to revise prior conversation turns through natural-language edit requests, treating the transcript as an editable representation of conversational state rather than an immutable…
C3M: Cross-Session Multimodal Memory Maintenance for Long-Horizon Tasks
The preprint proposes bounded query-blind memory for long-horizon tasks. It describes relation-aware updates for complementary and incompatible records, budgeted routing that selects index pages and expands source…
CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents
CliffCompaction is an autocompaction technique for long-horizon coding agents that reduces context cost while preserving or improving benchmark performance. It only truncates or drops content rather than rephrasing it…
CompKV: Compensation-Aware KV Selection for Long-Context LLM Inference
CompKV selects token blocks jointly with the downstream compensation mechanism for sparse attention rather than choosing tokens only by attention mass. The preprint derives a block-level residual approximation based on…
AutoViewMem: Self-Configuring Orthogonal Views for Conversational Long-Term Memory
AutoViewMem discovers low-overlap semantic views from interaction traces, uses a compact complementary view set for structured write-time extraction, and grounds memories in provenance. It moves semantic…
Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design
Designer-RSI uses a frozen frontier model operating professional design software through more than 230 tools and maintains an external procedural-memory bank of natural-language skills. The memory widens by adding…
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
The record describes a 552B-backbone multimodal mixture-of-experts model with up to one million tokens of context, activating 16B parameters during decode and 8B during prefill. It reports cross-layer KV reuse, FP4 KV…
MemTrapBench: Benchmarking Cognitive Traps in LLM Memory Use
Researchers have introduced a new benchmark called MemTrapBench to test the performance of large language models (LLMs) in terms of cognitive traps caused by their memory components. The study identifies two types of…
Adaptive Memory and Reflection Multi-Agent System for Medical Question Answering
Researchers have developed an adaptive memory and reflection multi-agent system for medical question answering. This system, which uses specialized agents with dedicated memory and feedback, aims to improve the…
rEDMRec: Distilling Large Language Model Reasoning into an Editable Experience Memory for Recommendation
The paper proposes a new approach for recommendation systems, called rEDMRec, which distills the reasoning of large language models into an editable memory that can be used to rank candidate items without regenerating…
EvoMem: Memory-Augmented Evolution for Code Optimization
The paper introduces EvoMem, a memory architecture for Long Learning Machine (LLM) based evolutionary program search. EvoMem captures and reuses knowledge from successful mutation events to improve the efficiency of…
ContextWeave: A Real-World Workflow Benchmark
Researchers introduce ContextWeave, a new benchmark for evaluating the performance of language agents in long-horizon, stateful workflows. The benchmark measures how well an agent can recall and apply past experiences…
ContextWeave: A Real-World Workflow Benchmark
The paper introduces ContextWeave, a new benchmark for evaluating the performance of language agents in long-horizon, stateful workflows. The benchmark focuses on memory as an essential component and measures workspace…
Hierarchical Graph Memory for LLM Agents with Path-level Localization and Rewrite
The authors propose HiGram, a hierarchical graph memory framework for long-term reasoning agents. This framework aims to efficiently update memories over time by organizing them into a coarse-to-fine architecture…
Hierarchical Graph Memory for LLM Agents with Path-level Localization and Rewrite
The authors propose HiGram, an evolving hierarchical graph memory framework that aims to improve the efficiency and effectiveness of long-term reasoning agents by reducing irrelevant information during retrieval and…
MemSecBench: Tracking Agent Memory Poisoning from Persistence to Consequence and Repair
Researchers have developed a benchmark called MemSecBench to evaluate the lifecycle security of agent memory systems. The benchmark contains 310 cases from various contexts, each following a specific protocol and using…
Eviction as Estimation: A Fixed-Lag Smoothing View of Test-Time Memory, and When Measuring Beats Accumulating
The authors present a new approach for managing a language model's working memory by treating the decision of which stored items to keep as an estimation problem. They introduce a method called 'fixed-lag smoothing'…
Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems
A research paper suggests a new approach to managing AI agents' reasoning context, which includes conversation histories, large prompts, and tool definitions. The authors argue that this should be treated as a…
Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems
The paper proposes a new approach for managing AI agent context, arguing that it's not just a storage-and-retrieval problem but a lifecycle issue involving deciding what to remember, structuring it, choosing the right…
Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems
The article proposes a new discipline called Agentic Context Management (ACM) to address issues faced by AI agents in managing their reasoning context, such as conversation histories and large prompts. The authors…
Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems
The paper discusses Agentic Context Management (ACM), a new approach to managing the context in which AI agents operate. It argues that current methods for managing an agent's context, such as storing and retrieving…
Agentic Context Management: Solving Agent Memory and Cost by Treating Them as Lifecycle and Architecture Problems
The paper presents Agentic Context Management (ACM), a solution to the issue of AI agents being overwhelmed by their own accumulating history and growing token cost in production settings. The authors argue that…
SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration
The paper introduces SearchOS, a system that aims to improve the efficiency of open-domain information seeking for information-seeking agents. It struggles with tracking task progress as interaction histories grow and…
Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents
The paper proposes a new framework called Memory as a Controlled Process (MemCon) for large language model (LLM) agents that use external memory systems. MemCon learns an online policy to adaptively decide when, what…
Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents
The paper presents a new framework called Memory as a Controlled Process (MemCon) for large language model (LLM) agents. MemCon aims to address the limitation of existing memory systems that rely on fixed…
LightMem-Ego: Your AI Memory for Everyday Life
The research paper introduces LightMem-Ego, a lightweight multimodal memory system designed for AI assistants on mobile and wearable devices. The system continuously captures and organizes egocentric visual and audio…
Cognitive-structured Multimodal Agent for Multimodal Understanding, Generation, and Editing
The paper proposes a new multimodal agent called the Cognitive-structured Multimodal Agent. This agent is designed to improve long-horizon multimodal dialogue by externalizing visual information into an Episodic Visual…
Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents
Researchers propose a proactive memory agent for long-horizon tasks in which an action agent must make decisions based on state information scattered across an expanding trajectory. The proposed memory agent updates a…
Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents
Researchers propose a proactive memory agent for long-horizon tasks where decision-relevant information can get buried as trajectories grow. The agent updates a structured memory bank from the recent trajectory and…
AutoMem: Automated Learning of Memory as a Cognitive Skill
Researchers developed AutoMem, a framework for improving a language model's memory management skills in long-horizon games. By treating memory as a trainable skill, they allowed the model to decide how to manage its…
AutoMem: Automated Learning of Memory as a Cognitive Skill
Researchers have developed a framework called AutoMem to improve the performance of large language models (LLMs) in long-horizon games by treating memory management as a learnable skill. They let the model decide how…
MemSyco-Bench: Benchmarking Sycophancy in Agent Memory
Researchers have proposed a new benchmark called MemSyco-Bench to evaluate how memory in large language models (LLMs) can lead to sycophancy, where agents over-align with the user at the cost of factual accuracy or…
Self-GC: Self-Governing Context for Long-Horizon LLM Agents
A research paper titled 'Self-GC: Self-Governing Context for Long-Horizon LLM Agents' presents a system called Self-GC that aims to manage the lifecycle of context objects for long-horizon language models (LLMs). The…
ACE: Pluggable Adaptive Context Elasticizer across Agents
Researchers propose a new module called Adaptive Context Elasticizer (ACE) to address the challenges posed by growing trajectory lengths in large language models (LLMs) with fixed context windows. ACE is designed to be…
ECHO: Prune to act, trace to learn with selective turn memory in agentic RL
Researchers propose ECHO, a selective turn-memory framework to address limitations of existing context management methods in long-horizon language agents. ECHO compresses completed environment turns into compact memory…
Advancing Omnimodal Embodied Agents from Isolated Skills to Everyday Physical Autonomy
The researchers present a framework called OmniAct that aims to build persistent embodied agents for operation in unstructured environments. The framework integrates a multimodal semantic planner, an adaptive…
