AlpsBench supplies 2500 real-dialogue sequences with verified memories to benchmark LLM extraction, updating, retrieval, and utilization of personalized information.
hub Canonical reference
What Deserves Memory: Adaptive Memory Distillation for LLM Agents
Canonical reference. 80% of citing Pith papers cite this work as background.
20
Pith papers citing it
Background
80% of classified citations
abstract
Memory systems for LLM agents struggle to determine what information deserves retention. Existing approaches rely on predefined heuristics such as importance scores, emotional tags, or factual templates, encoding designer intuition rather than learning from the data itself. Inspired by cognitive ideas, we propose NEMORI, an adaptive memory distillation framework that casts the assessment of experience's future utility as a matter of predictability. Specifically, NEMORI comprises two cascading modules: Episodic Memory Integration transforms raw interactions into coherent narratives, and Semantic Knowledge Distillation extracts insights via prediction error. Centering on distillation, the framework remains agnostic to downstream management. Extensive experiments confirm that NEMORI achieves strong performance, efficiency, and storage reduction. Our work suggests that observing the intrinsic properties of interaction sequences offers a viable, data-driven alternative to heuristic-based memory design. Code: https://github.com/nemori-ai/nemori.
hub tools
citation-role summary
background 4
baseline 1
citation-polarity summary
years
2026 20verdicts
UNVERDICTED 20representative citing papers
TRACE models conversations as hierarchical graphs with temporal, causal, update, and contradiction relations plus validity annotations to enable bounded, state-aware query processing over long conversational histories.
MAGE manages agent memory via a hierarchical state tree with Grow, Compress, Maintain, and Revise operations, raising success rates 7.8-20.4 pp and cutting tokens 55.1% on MemoryArena.
Memory for long-horizon agents should preserve distinctions that affect decisions under a fixed budget, not descriptive features, yielding an exact forgetting boundary and a new online learner DeMem with regret guarantees.
Memora benchmark and FAMA metric show that LLMs and memory agents frequently reuse invalid memories and struggle to reconcile evolving information in long-term interactions.
xMemory builds revisable hierarchical agent memory by segmenting histories, decoupling into components, and aggregating via sparsity-semantic objective, yielding better answer quality and lower token use than flat RAG on LoCoMo and PerLTQA.
FluxMem evolves memory as a heterogeneous graph via three refinement stages and reports consistent state-of-the-art results on LoCoMo, Mind2Web, and GAIA benchmarks.
Auto-Dreamer trains an offline memory consolidator via GRPO on agent performance to abstract cross-session patterns, outperforming baselines by 7 points on ScienceWorld with 12x smaller memory and generalizing to ALFWorld and WebArena.
HAGE proposes a trainable weighted graph memory framework with LLM intent classification, dynamic edge modulation, and RL optimization that improves long-horizon reasoning accuracy in agentic LLMs over static baselines.
The paper defines and evaluates Trojan Hippo attacks on LLM agent memory, showing 85-100% success in data exfiltration across backends and reduced rates with defenses at varying utility costs.
A lightweight supervised router using frozen-LLM embeddings for memory admission decisions outperforms LLM-based memory managers in both F1 score and latency on the LoCoMo benchmark.
APEX-MEM uses property graphs with temporal events, append-only storage, and an agentic retrieval system to reach 88.88% accuracy on LOCOMO QA and 86.2% on LongMemEval, outperforming prior session-aware methods.
TSUBASA improves long-horizon personalization in LLMs via dynamic memory evolution for writing and context-distillation self-learning for reading, outperforming Mem0 and Memory-R1 on Qwen-3 benchmarks while reducing token use.
Opal enables private long-term memory for personal AI by decoupling reasoning to a trusted enclave with a lightweight knowledge graph and piggybacking reindexing on ORAM accesses.
HyMem introduces dual-granular memory storage with a lightweight summary module for fast responses and selective activation of a deep LLM module for complex queries, outperforming full-context baselines by 92.6% lower computational cost on LOCOMO and LongMemEval benchmarks.
LRE is a kilobyte-scale learned relevance eviction policy for LLM agent memory that matches or exceeds full-history accuracy while cutting peak context size by up to 52% and using zero compressor calls.
HyperMem is a hypergraph memory architecture that groups related conversation episodes and facts via hyperedges and reports 92.73% LLM-as-a-judge accuracy on the LoCoMo benchmark.
MemCoT transforms long-context LLM reasoning into an iterative stateful search using multi-view memory for evidence localization and dual short-term memory for guiding decisions, achieving SOTA on LoCoMo and LongMemEval-S benchmarks.
LLM agent memory is organized into Storage (preserving trajectories), Reflection (refining them), and Experience (abstracting into reusable knowledge) stages driven by needs for long-range consistency, dynamic adaptation, and continual learning.
A survey that defines agent skills as reusable procedural artifacts and reviews methods, resources, and applications across their representation, acquisition, retrieval, and evolution stages.
citing papers explorer
-
AlpsBench: An LLM Personalization Benchmark for Real-Dialogue Memorization and Preference Alignment
AlpsBench supplies 2500 real-dialogue sequences with verified memories to benchmark LLM extraction, updating, retrieval, and utilization of personalized information.
-
TRACE: State-Aware Query Processing over Temporal Evidence Graphs for Conversational Data
TRACE models conversations as hierarchical graphs with temporal, causal, update, and contradiction relations plus validity annotations to enable bounded, state-aware query processing over long conversational histories.
-
Beyond Semantic Organization: Memory as Execution State Management for Long-Horizon Agents
MAGE manages agent memory via a hierarchical state tree with Grow, Compress, Maintain, and Revise operations, raising success rates 7.8-20.4 pp and cutting tokens 55.1% on MemoryArena.
-
Remember the Decision, Not the Description: A Rate-Distortion Framework for Agent Memory
Memory for long-horizon agents should preserve distinctions that affect decisions under a fixed budget, not descriptive features, yielding an exact forgetting boundary and a new online learner DeMem with regret guarantees.
-
From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents
Memora benchmark and FAMA metric show that LLMs and memory agents frequently reuse invalid memories and struggle to reconcile evolving information in long-term interactions.
-
Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation
xMemory builds revisable hierarchical agent memory by segmenting histories, decoupling into components, and aggregating via sparsity-semantic objective, yielding better answer quality and lower token use than flat RAG on LoCoMo and PerLTQA.
-
Rethinking Memory as Continuously Evolving Connectivity
FluxMem evolves memory as a heterogeneous graph via three refinement stages and reports consistent state-of-the-art results on LoCoMo, Mind2Web, and GAIA benchmarks.
-
Auto-Dreamer: Learning Offline Memory Consolidation for Language Agents
Auto-Dreamer trains an offline memory consolidator via GRPO on agent performance to abstract cross-session patterns, outperforming baselines by 7 points on ScienceWorld with 12x smaller memory and generalizing to ALFWorld and WebArena.
-
HAGE: Harnessing Agentic Memory via RL-Driven Weighted Graph Evolution
HAGE proposes a trainable weighted graph memory framework with LLM intent classification, dynamic edge modulation, and RL optimization that improves long-horizon reasoning accuracy in agentic LLMs over static baselines.
-
Trojan Hippo: Weaponizing Agent Memory for Data Exfiltration
The paper defines and evaluates Trojan Hippo attacks on LLM agent memory, showing 85-100% success in data exfiltration across backends and reduced rates with defenses at varying utility costs.
-
MemRouter: Memory-as-Embedding Routing for Long-Term Conversational Agents
A lightweight supervised router using frozen-LLM embeddings for memory admission decisions outperforms LLM-based memory managers in both F1 score and latency on the LoCoMo benchmark.
-
APEX-MEM: Agentic Semi-Structured Memory with Temporal Reasoning for Long-Term Conversational AI
APEX-MEM uses property graphs with temporal events, append-only storage, and an agentic retrieval system to reach 88.88% accuracy on LOCOMO QA and 86.2% on LongMemEval, outperforming prior session-aware methods.
-
TSUBASA: Improving Long-Horizon Personalization via Evolving Memory and Self-Learning with Context Distillation
TSUBASA improves long-horizon personalization in LLMs via dynamic memory evolution for writing and context-distillation self-learning for reading, outperforming Mem0 and Memory-R1 on Qwen-3 benchmarks while reducing token use.
-
Opal: Private Memory for Personal AI
Opal enables private long-term memory for personal AI by decoupling reasoning to a trusted enclave with a lightweight knowledge graph and piggybacking reindexing on ORAM accesses.
-
HyMem: Hybrid Memory Architecture with Dynamic Retrieval Scheduling
HyMem introduces dual-granular memory storage with a lightweight summary module for fast responses and selective activation of a deep LLM module for complex queries, outperforming full-context baselines by 92.6% lower computational cost on LOCOMO and LongMemEval benchmarks.
-
Learning What Not to Forget: Long-Horizon Agent Memory from a Few Kilobytes of Learning
LRE is a kilobyte-scale learned relevance eviction policy for LLM agent memory that matches or exceeds full-history accuracy while cutting peak context size by up to 52% and using zero compressor calls.
-
HyperMem: Hypergraph Memory for Long-Term Conversations
HyperMem is a hypergraph memory architecture that groups related conversation episodes and facts via hyperedges and reports 92.73% LLM-as-a-judge accuracy on the LoCoMo benchmark.
-
MemCoT: Test-Time Scaling through Memory-Driven Chain-of-Thought
MemCoT transforms long-context LLM reasoning into an iterative stateful search using multi-view memory for evidence localization and dual short-term memory for guiding decisions, achieving SOTA on LoCoMo and LongMemEval-S benchmarks.
-
From Storage to Experience: A Survey on the Evolution of LLM Agent Memory Mechanisms
LLM agent memory is organized into Storage (preserving trajectories), Reflection (refining them), and Experience (abstracting into reusable knowledge) stages driven by needs for long-range consistency, dynamic adaptation, and continual learning.
-
A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications
A survey that defines agent skills as reusable procedural artifacts and reviews methods, resources, and applications across their representation, acquisition, retrieval, and evolution stages.